Merge pull request #6353 from feeeei/main

refactor(billing): 长上下文阶梯计价改为价格目录数据驱动

.dockerignore

@@ -49,6 +49,9 @@ coverage/
 .env.*
 !.env.example
 
+# 本地闭源插件目录可能包含发布签名私钥，绝不能进入 Docker 构建上下文。
+/plugins/
+
 # Local config
 config.yaml
 config.local.yaml

.github/audit-exceptions.yml

@@ -35,3 +35,10 @@ exceptions:
     mitigation: "Proxy configuration not user-controlled; upgrade when axios releases fix"
     expires_on: "2026-07-10"
     owner: "security@your-domain"
+  - package: nanoid
+    advisory: "GHSA-2v37-7h3g-55p8"
+    severity: high
+    reason: "Custom generator with size=0 not used; all nanoid calls use default size or positive values"
+    mitigation: "No zero-size custom generators in codebase; upgrade when nanoid releases fix"
+    expires_on: "2026-10-13"
+    owner: "security@your-domain"

.github/workflows/backend-ci.yml

@@ -17,6 +17,7 @@ jobs:
           /bin/bash -n deploy/apple-container.sh
           /bin/bash deploy/tests/apple-container-test.sh
           /bin/sh deploy/tests/docker-compose-security-test.sh
+          /bin/sh deploy/tests/docker-compose-gateway-env-test.sh
           /bin/sh deploy/tests/docker-runtime-resources-test.sh
           /bin/sh deploy/test-caddyfile-cache.sh
 
@@ -32,7 +33,7 @@ jobs:
           cache-dependency-path: backend/go.sum
       - name: Verify Go version
         run: |
-          go version | grep -q 'go1.26.6'
+          go version | grep -q 'go1.27.0'
       - name: Unit tests
         working-directory: backend
         run: make test-unit
@@ -72,10 +73,10 @@ jobs:
           cache-dependency-path: backend/go.sum
       - name: Verify Go version
         run: |
-          go version | grep -q 'go1.26.6'
+          go version | grep -q 'go1.27.0'
       - name: golangci-lint
         uses: golangci/golangci-lint-action@v9
         with:
-          version: v2.9
+          version: v2.13
           args: --timeout=30m
           working-directory: backend

.github/workflows/release.yml

@@ -115,7 +115,7 @@ jobs:
 
       - name: Verify Go version
         run: |
-          go version | grep -q 'go1.26.6'
+          go version | grep -q 'go1.27.0'
 
       # Docker setup for GoReleaser
       - name: Set up QEMU

.github/workflows/security-scan.yml

@@ -23,7 +23,7 @@ jobs:
           cache-dependency-path: backend/go.sum
       - name: Verify Go version
         run: |
-          go version | grep -q 'go1.26.6'
+          go version | grep -q 'go1.27.0'
       - name: Run govulncheck
         working-directory: backend
         run: |

.gitignore

@@ -1,4 +1,5 @@
-docs/claude-relay-service/
+# 本地杂项文档、测试数据和外部项目副本
+/docs-local/
 .codex
 
 # ===================
@@ -136,13 +137,23 @@ docs/*
 !docs/PAYMENT_CN.md
 !docs/ADMIN_PAYMENT_INTEGRATION_API.md
 !docs/ASYNC_IMAGE_TASKS.md
+!docs/BATCH_IMAGE_MVP.md
+!docs/COMPOSITE_GROUPS.md
+!docs/PLUGIN_DEVELOPMENT.md
+!docs/channel-monitor-v2-safe-defaults.md
 !docs/legal/
 !docs/legal/*.md
+!docs/screenshots/
+docs/screenshots/*
+!docs/screenshots/mobile-account-actions-menu.png
 .serena/
 .codex/
 frontend/coverage/
 aicodex
 output/
+.codegraph/
 
 # Vitest / Vite cache at repo root
 .vite/
+# 本地闭源插件开发目录及构建产物不进入 Sub2API 仓库
+/plugins/

DEV_GUIDE.md

@@ -34,8 +34,8 @@
 ### 开发工具
 
 ```bash
-# golangci-lint（CI 用 v2.9，本地建议装同一版以免版本差异带来的噪音）
-go install github.com/golangci/golangci-lint/v2/cmd/golangci-lint@v2.9
+# golangci-lint（CI 用 v2.13，本地建议装同一版以免版本差异带来的噪音）
+go install github.com/golangci/golangci-lint/v2/cmd/golangci-lint@v2.13
 
 # pnpm (前端包管理)
 npm install -g pnpm
@@ -47,13 +47,13 @@ npm install -g pnpm
 
 | Workflow | 触发条件 | 检查内容 |
 |----------|----------|----------|
-| **backend-ci.yml** | push, pull_request | 单元测试 + 集成测试 + golangci-lint v2.9 |
+| **backend-ci.yml** | push, pull_request | 单元测试 + 集成测试 + golangci-lint v2.13 |
 | **security-scan.yml** | push, pull_request, 每周一 | govulncheck + gosec + pnpm audit |
 | **release.yml** | tag `v*` | 构建发布（PR 不触发） |
 
 ### CI 要求
 
-- Go 版本必须是 **1.26.6**：三个 workflow 都用 `go-version-file: backend/go.mod` 取版本，随后硬断言 `go version | grep -q 'go1.26.6'`。升级 Go 时要同时改 `backend/go.mod`、`backend-ci.yml`（两处）、`release.yml`、`security-scan.yml` 里的这句断言，**以及三个 Dockerfile 里的 Go 构建镜像**（`Dockerfile` / `deploy/Dockerfile` 的 `ARG GOLANG_IMAGE`、`backend/Dockerfile` 的 `FROM golang:`）。前者漏了 CI 会在版本校验步骤直接失败；**后者漏了 CI 不会报，而是等到有人用这些 Dockerfile 构建时才失败**（`go.mod requires go >= X (running Y; GOTOOLCHAIN=local)`）。
+- Go 版本必须是 **1.27.0**：三个 workflow 都用 `go-version-file: backend/go.mod` 取版本，随后硬断言 `go version | grep -q 'go1.27.0'`。升级 Go 时要同时改 `backend/go.mod`、`backend-ci.yml`（两处）、`release.yml`、`security-scan.yml` 里的这句断言，**以及三个 Dockerfile 里的 Go 构建镜像**（`Dockerfile` / `deploy/Dockerfile` 的 `ARG GOLANG_IMAGE`、`backend/Dockerfile` 的 `FROM golang:`）。前者漏了 CI 会在版本校验步骤直接失败；**后者漏了 CI 不会报，而是等到有人用这些 Dockerfile 构建时才失败**（`go.mod requires go >= X (running Y; GOTOOLCHAIN=local)`）。
 - 前端使用 `pnpm install --frozen-lockfile`，必须提交 `pnpm-lock.yaml`
 
 ### 本地测试命令
@@ -203,7 +203,7 @@ go test -tags=integration ./...
 **解决**：
 ```bash
 cd backend
-go generate ./ent  # 重新生成 ent 代码
+go generate ./ent  # 重新生成 ent 代码（json.RawMessage 字段会生成为同类型的 jsontext.Value，属预期）
 git add ent/       # 生成的文件也要提交
 ```
 

Dockerfile

@@ -8,7 +8,7 @@
 # =============================================================================
 
 ARG NODE_IMAGE=node:24-alpine
-ARG GOLANG_IMAGE=golang:1.26.6-alpine
+ARG GOLANG_IMAGE=golang:1.27.0-alpine
 ARG ALPINE_IMAGE=alpine:3.21
 ARG POSTGRES_IMAGE=postgres:18-alpine
 ARG GOPROXY=https://goproxy.cn,direct

README.md

@@ -4,7 +4,7 @@
 
 # Sub2API
 
-[![Go](https://img.shields.io/badge/Go-1.26.5-00ADD8.svg)](https://golang.org/)
+[![Go](https://img.shields.io/badge/Go-1.27.0-00ADD8.svg)](https://golang.org/)
 [![Vue](https://img.shields.io/badge/Vue-3.4+-4FC08D.svg)](https://vuejs.org/)
 [![PostgreSQL](https://img.shields.io/badge/PostgreSQL-15+-336791.svg)](https://www.postgresql.org/)
 [![Redis](https://img.shields.io/badge/Redis-7+-DC382D.svg)](https://redis.io/)
@@ -58,6 +58,11 @@ Please read the following carefully before using this project:
 <td>Thanks to AIGoCode for sponsoring this project! AIGoCode is an all-in-one platform that integrates Claude Code, Codex, and the latest Gemini models, providing you with stable, efficient, and highly cost-effective AI coding services. The platform offers flexible subscription plans, zero risk of account suspension, direct access with no VPN required, and lightning-fast responses. AIGoCode has prepared a special benefit for sub2api users: if you register via <a href="https://aigocode.com/invite/SUB2API">this link</a>, you'll receive an extra 10% bonus credit on your first top-up!</td>
 </tr>
 
+<tr>
+<td width="180"><a href="https://codex-everywhere.com"><img src="assets/partners/logos/codex-everywhere.jpg" alt="CodexEverywhere" width="150"></a></td>
+<td>Real GPT-5.6 series at 3% of OpenAI pricing — <a href="https://codex-everywhere.com">CodexEverywhere</a> is democratizing access to frontier models for developers worldwide. We believe in transparency and honesty, with model quality verified by active community oversight for months. USD and crypto friendly. Start with a free $20 trial at <a href="https://codex-everywhere.com">codex-everywhere.com</a>.</td>
+</tr>
+
 <tr>
 <td width="180"><a href="https://shop.bmoplus.com/?utm_source=github"><img src="assets/partners/logos/bmoplus.jpg" alt="bmoplus" width="150"></a></td>
 <td>Huge thanks to BmoPlus for sponsoring this project! BmoPlus is a highly reliable AI account provider built strictly for heavy AI users and developers. They offer rock-solid, ready-to-use accounts and official top-up services for ChatGPT Plus / ChatGPT Pro (Full Warranty) / Claude Pro / Super Grok / Gemini Pro. By registering and ordering through <a href="https://shop.bmoplus.com/?utm_source=github">BmoPlus - Premium AI Accounts & Top-ups</a>, users can unlock the mind-blowing rate of 10% of the official GPT subscription price (90% OFF)</td>
@@ -117,18 +122,6 @@ Please read the following carefully before using this project:
 </td>
 </tr>
 
-<tr>
-<td width="180"><a href="https://console.claudeapi.com/agent/register/drTKjyn6wGLK061Z"><img src="assets/partners/logos/claudeapi.jpg" alt="claudeapi" width="150"></a></td>
-<td>Thanks to Claude API for sponsoring this project! <a href="https://console.claudeapi.com/agent/register/drTKjyn6wGLK061Z">Claude API</a> is an official-channel API provider focused on Claude models. Built on official Anthropic keys and the official AWS Bedrock channel, it delivers a stable integration experience for Claude Code and Agent applications, supports the full Claude model lineup, and retains official capabilities such as Tool Use and long context. The service involves no reverse engineering and no model degradation, making it a great fit for heavy Claude Code users, Agent engineers, and enterprise engineering teams. Register via the <a href="https://console.claudeapi.com/agent/register/drTKjyn6wGLK061Z">exclusive link</a> and contact customer support to receive free trial credits; invoicing and team onboarding are also supported.
-</td>
-</tr>
-
-<tr>
-<td width="180"><a href="https://code0.ai/agent/register/LgpIgl9JHtVG53V1?utm_source=zcf&utm_medium=partner&utm_campaign=zcf_2026&utm_content=default"><img src="assets/partners/logos/code0.jpg" alt="code0" width="150"></a></td>
-<td>Thanks to code0.ai for sponsoring this project! <a href="https://code0.ai/agent/register/LgpIgl9JHtVG53V1?utm_source=zcf&utm_medium=partner&utm_campaign=zcf_2026&utm_content=default">code0.ai</a> is an AI coding workbench for developers and engineering teams, aggregating mainstream agent coding capabilities such as Claude Code and Codex, and covering common development scenarios including code generation, project understanding, debugging and fixing, code review, and documentation generation. It suits independent developers, Agent engineers, open-source maintainers, and enterprise R&D teams, with invoicing and team onboarding supported. Register via the <a href="https://code0.ai/agent/register/LgpIgl9JHtVG53V1?utm_source=zcf&utm_medium=partner&utm_campaign=zcf_2026&utm_content=default">exclusive link</a> and contact customer support to receive free trial credits and experience a more efficient AI coding workflow.
-</td>
-</tr>
-
 <tr>
 <td width="180"><a href="https://nagora.ai/"><img src="assets/partners/logos/nagora.png" alt="Nagora" width="150"></a></td>
 <td><a href="https://nagora.ai/">Nagora</a> is a multi-model AI API gateway built for developers and teams. With a single account and API key, you can access more than 26 leading text and image models through one unified interface. It is compatible with OpenAI, Anthropic, and Gemini protocols and integrates seamlessly with development tools such as Claude Code, Codex, and Gemini CLI. The platform provides intelligent routing, automatic failover, transparent pricing, and consolidated billing, along with budget management, rate limiting, and concurrency controls. This makes AI usage more reliable and manageable across individual development, team collaboration, and production environments. No changes to your existing application are required. Simply replace the Base URL and API key to complete the integration in as little as one minute.</td>
@@ -174,6 +167,16 @@ Please read the following carefully before using this project:
 <td><a href="https://www.duckip.cn/?keyword=cu7oog6y">DuckIP</a> - 90M+ global residential network resources across 195+ countries and regions, with rotation and sticky sessions for public data collection, RAG updates, model evaluation, and multi-region data workloads. 🟢Residential Proxy - 20% Off; 🟢Static Residential Proxy - Starting at ¥50.00/IP; 🟢Unlimited Residential Proxy - Starting at ¥19.8/Hour. ✅Get 500M Free Trial.</td>
 </tr>
 
+<tr>
+<td width="180"><a href="https://go.apimart.ai/gh-sub2api"><img src="assets/partners/logos/apimart.jpg" alt="APIMart" width="150"></a></td>
+<td>Thanks to APIMart for sponsoring this project! <a href="https://go.apimart.ai/gh-sub2api">APIMart</a> is a low-cost API platform for AI image and video generation — GPT-Image-2 from $0.006 per image, with 160+ images per dollar. One async API covers both image and video: submit a task, get an ID, and retrieve results via polling or callback. Batch tens of thousands of images without timeouts, and switch models without changing code. Pay as you go with no monthly fee — <a href="https://go.apimart.ai/gh-sub2api">sign up here</a> to get started.</td>
+</tr>
+
+<tr>
+<td width="180"><a href="https://www.axisnow.io/"><img src="assets/partners/logos/axisnow.jpg" alt="AxisNow" width="150"></a></td>
+<td>Thanks to AxisNow for sponsoring this project! <a href="https://www.axisnow.io/">AxisNow</a> is building a cloud-agnostic edge control layer. Use AxisNow to protect and accelerate your websites and APIs, delivering optimal access experiences for both mainland China and global users, while extending acceleration and security capabilities to native/mobile apps through client SDKs — all on your own controllable, composable edge network.</td>
+</tr>
+
 </table>
 
 ## Overview
@@ -206,7 +209,7 @@ Community projects that extend or integrate with Sub2API:
 
 | Component | Technology |
 |-----------|------------|
-| Backend | Go 1.26.5, Gin, Ent |
+| Backend | Go 1.27.0, Gin, Ent |
 | Frontend | Vue 3.4+, Vite 5+, TailwindCSS |
 | Database | PostgreSQL 15+ |
 | Cache/Queue | Redis 7+ |
@@ -550,12 +553,6 @@ default:
   rate_multiplier: 1.0
 ```
 
-### Sora Status (Temporarily Unavailable)
-
-> ⚠️ Sora-related features are temporarily unavailable due to technical issues in upstream integration and media delivery.
-> Please do not rely on Sora in production at this time.
-> Existing `gateway.sora_*` configuration keys are reserved and may not take effect until these issues are resolved.
-
 Additional security-related options are available in `config.yaml`:
 
 - `cors.allowed_origins` for CORS allowlist
@@ -658,6 +655,32 @@ Or set `GATEWAY_OPENAI_WS_MODE_ROUTER_V2_ENABLED=true` in the environment.
 Use `http_bridge` for client-WebSocket/upstream-HTTP operation when rolling out
 or mitigating upstream WebSocket issues.
 
+#### Force OpenAI upstream HTTP/SSE
+
+When an egress proxy or network repeatedly reconnects OpenAI Responses
+WebSockets, set the global fallback in the persisted deployment configuration:
+
+```yaml
+gateway:
+  openai_ws:
+    force_http: true
+```
+
+For Compose and Apple container deployments, the equivalent `.env` setting is:
+
+```bash
+GATEWAY_OPENAI_WS_FORCE_HTTP=true
+```
+
+This selects HTTP/SSE for OpenAI upstream Responses traffic that would
+otherwise use WebSocket. It does not change the client-facing protocol or force
+HTTP/1.1; configure `gateway.openai_http2.enabled` (or
+`GATEWAY_OPENAI_HTTP2_ENABLED=false`) separately when a proxy is incompatible
+with HTTP/2. Unlike the account-level `http_bridge` mode, this global fallback
+takes effect without enabling `mode_router_v2_enabled`. Keep the setting in the
+deployment's persisted `.env` or `config.yaml`, rather than inside a running
+container, so it is read again after an image update or container recreation.
+
 #### ⚠️ Important: Creating the Admin Account
 
 The initial admin account is **only created via the setup wizard** (served at `http://<host>:8080` on first run). The `default.admin_email` / `default.admin_password` fields in `config.yaml` are **not used** to create it — they exist in the template for historical reasons.
@@ -736,7 +759,7 @@ Sub2API supports both Grok subscription accounts through xAI OAuth and standard
 - Codex CLI style Responses WebSocket ingress is accepted on the Responses targets and bridged to xAI HTTP/SSE Responses upstream
 - Text models: `grok-4.5`, `grok-4.3`, `grok-build-0.1`, `grok-composer-2.5-fast`, `grok-4.20-0309-reasoning`, `grok-4.20-0309-non-reasoning`, and `grok-4.20-multi-agent-0309`
 - Media targets for Grok groups: `/v1/images/generations`, `/images/generations`, `/v1/images/edits`, `/images/edits`, `/v1/videos/generations`, `/videos/generations`, `/v1/videos/edits`, `/videos/edits`, `/v1/videos/extensions`, `/videos/extensions`, `/v1/videos/{request_id}`, and `/videos/{request_id}`. Generation, editing, and extension requests require the group image-generation permission.
-- Media models: `grok-imagine`, `grok-imagine-image-quality`, `grok-imagine-image`, `grok-imagine-edit`, `grok-imagine-video`, and `grok-imagine-video-1.5`
+- Media models: `grok-imagine`, `grok-imagine-image-quality`, `grok-imagine-image`, `grok-imagine-image-2.0`, `grok-imagine-edit`, `grok-imagine-video`, and `grok-imagine-video-1.5`
 - JSON image-edit and video-generation requests accept image references in `image`, `images`, `reference_images`, and `mask` objects. Use `url` for xAI-compatible payloads; the legacy `image_url` field remains accepted and is normalized to `url` before forwarding.
 - Out of scope for this provider: TTS, transcription, browser automation, cookies, and Grok web scraping
 
@@ -866,11 +889,11 @@ sub2api/
 
 ## Star History
 
-<a href="https://star-history.com/#Wei-Shaw/sub2api&Date">
+<a href="https://star-history.dera.page/#Wei-Shaw/sub2api&Date">
  <picture>
-   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=Wei-Shaw/sub2api&type=Date&theme=dark" />
-   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=Wei-Shaw/sub2api&type=Date" />
-   <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=Wei-Shaw/sub2api&type=Date" />
+   <source media="(prefers-color-scheme: dark)" srcset="https://star-history.dera.page/svg?repos=Wei-Shaw/sub2api&type=Date&theme=dark" />
+   <source media="(prefers-color-scheme: light)" srcset="https://star-history.dera.page/svg?repos=Wei-Shaw/sub2api&type=Date" />
+   <img alt="Star History Chart" src="https://star-history.dera.page/svg?repos=Wei-Shaw/sub2api&type=Date" />
  </picture>
 </a>
 

README_CN.md

@@ -4,7 +4,7 @@
 
 # Sub2API
 
-[![Go](https://img.shields.io/badge/Go-1.26.5-00ADD8.svg)](https://golang.org/)
+[![Go](https://img.shields.io/badge/Go-1.27.0-00ADD8.svg)](https://golang.org/)
 [![Vue](https://img.shields.io/badge/Vue-3.4+-4FC08D.svg)](https://vuejs.org/)
 [![PostgreSQL](https://img.shields.io/badge/PostgreSQL-15+-336791.svg)](https://www.postgresql.org/)
 [![Redis](https://img.shields.io/badge/Redis-7+-DC382D.svg)](https://redis.io/)
@@ -59,6 +59,11 @@
 <td>感谢 AIGoCode 赞助了本项目！AIGoCode 是一站式集成 Claude Code、Codex 以及最新 Gemini 模型的综合平台，为您提供稳定、高效、高性价比的 AI 编程服务。平台提供灵活的订阅方案，零封号风险，免 VPN 直连，响应极速。AIGoCode 为 sub2api 用户准备了专属福利：通过<a href="https://aigocode.com/invite/SUB2API">此链接</a>注册，首次充值可额外获得 10% 赠送额度！</td>
 </tr>
 
+<tr>
+<td width="180"><a href="https://codex-everywhere.com"><img src="assets/partners/logos/codex-everywhere.jpg" alt="CodexEverywhere" width="150"></a></td>
+<td>Real GPT-5.6 series at 3% of OpenAI pricing — <a href="https://codex-everywhere.com">CodexEverywhere</a> is democratizing access to frontier models for developers worldwide. We believe in transparency and honesty, with model quality verified by active community oversight for months. USD and crypto friendly. Start with a free $20 trial at <a href="https://codex-everywhere.com">codex-everywhere.com</a>.</td>
+</tr>
+
 <tr>
 <td width="180"><a href="https://shop.bmoplus.com/?utm_source=github"><img src="assets/partners/logos/bmoplus.jpg" alt="bmoplus" width="150"></a></td>
 <td>感谢 BmoPlus 赞助了本项目！BmoPlus 是一家专为AI订阅重度用户打造的可靠 AI 账号代充服务商，提供稳定的 ChatGPT Plus / ChatGPT Pro(全程质保) / Claude Pro / Super Grok / Gemini Pro 的官方代充&成品账号。 通过<a href="https://shop.bmoplus.com/?utm_source=github">BmoPlus AI成品号专卖/代充</a>注册下单的用户，可享GPT 官网订阅一折 的震撼价格！</td>
@@ -120,18 +125,6 @@
 </td>
 </tr>
 
-<tr>
-<td width="180"><a href="https://console.claudeapi.com/agent/register/drTKjyn6wGLK061Z"><img src="assets/partners/logos/claudeapi.jpg" alt="claudeapi" width="150"></a></td>
-<td>感谢 Claude API 对本项目的赞助！ <a href="https://console.claudeapi.com/agent/register/drTKjyn6wGLK061Z">Claude API</a> 是专注 Claude 模型的官方渠道 API 服务商，基于 Anthropic 官方 Key 与 AWS Bedrock 官方渠道，提供稳定的 Claude Code 与 Agent 应用接入体验，支持 Claude 全系列模型，保留 Tool Use、长上下文等官方能力。服务非逆向、非降智，适合 Claude Code 深度用户、Agent 工程师与企业技术团队使用。通过<a href="https://console.claudeapi.com/agent/register/drTKjyn6wGLK061Z">[专属链接]</a>注册后联系客服，可领取免费测试额度，并支持开票和团队对接。
-</td>
-</tr>
-
-<tr>
-<td width="180"><a href="https://code0.ai/agent/register/LgpIgl9JHtVG53V1?utm_source=zcf&utm_medium=partner&utm_campaign=zcf_2026&utm_content=default"><img src="assets/partners/logos/code0.jpg" alt="code0" width="150"></a></td>
-<td>感谢 code0.ai 对本项目的赞助！ <a href="https://code0.ai/agent/register/LgpIgl9JHtVG53V1?utm_source=zcf&utm_medium=partner&utm_campaign=zcf_2026&utm_content=default">code0.ai</a> 是面向开发者与技术团队的 AI 编程工作台，聚合 Claude Code、Codex 等主流 Agent 编程能力，支持代码生成、项目理解、调试修复、代码审查与文档生成等常见研发场景。适合独立开发者、Agent 工程师、开源项目维护者和企业研发团队使用，支持开票和团队对接。通过<a href="https://code0.ai/agent/register/LgpIgl9JHtVG53V1?utm_source=zcf&utm_medium=partner&utm_campaign=zcf_2026&utm_content=default">[专属链接]</a>注册后联系客服，可领取免费测试额度，体验更高效的 AI 编程工作流。
-</td>
-</tr>
-
 <tr>
 <td width="180"><a href="https://nagora.ai/"><img src="assets/partners/logos/nagora.png" alt="Nagora" width="150"></a></td>
 <td><a href="https://nagora.ai/">Nagora</a> 是专为开发者和团队打造的多模型 AI API 网关。通过一个账户和一枚 API Key，即可统一调用 26+ 款主流文本与图像模型，兼容 OpenAI、Anthropic 与 Gemini 协议，并可无缝接入 Claude Code、Codex、Gemini CLI 等开发工具。平台提供智能路由、自动故障转移、透明计费与统一账单，同时支持预算、限速、并发控制，让个人开发、团队协作和生产环境中的 AI 调用更稳定、更可控。无需改造现有应用，只需替换 Base URL 与 API Key，最快 1 分钟即可完成接入。</td>
@@ -177,6 +170,16 @@
 <td><a href="https://www.duckip.cn/?keyword=cu7oog6y">DuckIP</a> - 9000 万+ 全球住宅网络资源，覆盖 195+ 国家和地区，支持轮换和粘性会话，适用于公共数据采集、RAG 更新、模型评估和多区域数据工作负载。🟢住宅代理 - 8 折优惠；🟢静态住宅代理 - ¥50.00/IP 起；🟢无限住宅代理 - ¥19.8/小时 起。✅免费领取 500M 试用流量。</td>
 </tr>
 
+<tr>
+<td width="180"><a href="https://go.apimart.ai/gh-sub2api"><img src="assets/partners/logos/apimart.jpg" alt="APIMart" width="150"></a></td>
+<td>感谢 APIMart 赞助了本项目！<a href="https://go.apimart.ai/gh-sub2api">APIMart</a> 是专注于 AI 图片/视频生成的低价 API 平台，GPT-Image-2 低至 $0.006/张，1 美元可生成 160+ 张图片。图片、视频一套异步 API 通吃：提交任务获取 ID，通过轮询或回调获取结果；批量生成上万张图片也不会超时，切换模型无需修改代码。按量付费、无月费，通过<a href="https://go.apimart.ai/gh-sub2api">此注册链接</a>注册即可开始使用。</td>
+</tr>
+
+<tr>
+<td width="180"><a href="https://www.axisnow.io/"><img src="assets/partners/logos/axisnow.jpg" alt="AxisNow" width="150"></a></td>
+<td>感谢 AxisNow 赞助了本项目！<a href="https://www.axisnow.io/">AxisNow</a> 正在打造一个云无关的边缘控制层。使用 AxisNow 可以保护并加速网站与 API，兼顾中国大陆及全球的访问体验，并通过客户端 SDK，将加速与安全能力延伸至原生/移动 App — 这些都在您可控的、自由组合的边缘网络。</td>
+</tr>
+
 </table>
 
 ## 项目概述
@@ -208,7 +211,7 @@ Sub2API 是一个 AI API 网关平台，用于分发和管理 AI 产品订阅的
 
 | 组件 | 技术 |
 |------|------|
-| 后端 | Go 1.26.5, Gin, Ent |
+| 后端 | Go 1.27.0, Gin, Ent |
 | 前端 | Vue 3.4+, Vite 5+, TailwindCSS |
 | 数据库 | PostgreSQL 15+ |
 | 缓存/队列 | Redis 7+ |
@@ -563,33 +566,6 @@ default:
   rate_multiplier: 1.0
 ```
 
-### Sora 功能状态（暂不可用）
-
-> ⚠️ 当前 Sora 相关功能因上游接入与媒体链路存在技术问题，暂时不可用。
-> 现阶段请勿在生产环境依赖 Sora 能力。
-> 文档中的 `gateway.sora_*` 配置仅作预留，待技术问题修复后再恢复可用。
-
-### Sora 媒体签名 URL（功能恢复后可选）
-
-当配置 `gateway.sora_media_signing_key` 且 `gateway.sora_media_signed_url_ttl_seconds > 0` 时，网关会将 Sora 输出的媒体地址改写为临时签名 URL（`/sora/media-signed/...`）。这样无需 API Key 即可在浏览器中直接访问，且具备过期控制与防篡改能力（签名包含 path + query）。
-
-```yaml
-gateway:
-  # /sora/media 是否强制要求 API Key（默认 false）
-  sora_media_require_api_key: false
-  # 媒体临时签名密钥（为空则禁用签名）
-  sora_media_signing_key: "your-signing-key"
-  # 临时签名 URL 有效期（秒）
-  sora_media_signed_url_ttl_seconds: 900
-```
-
-> 若未配置签名密钥，`/sora/media-signed` 将返回 503。  
-> 如需更严格的访问控制，可将 `sora_media_require_api_key` 设为 true，仅允许携带 API Key 的 `/sora/media` 访问。
-
-访问策略说明：
-- `/sora/media`：内部调用或客户端携带 API Key 才能下载
-- `/sora/media-signed`：外部可访问，但有签名 + 过期控制
-
 `config.yaml` 还支持以下安全相关配置：
 
 - `cors.allowed_origins` 配置 CORS 白名单
@@ -797,11 +773,11 @@ sub2api/
 
 ## Star History
 
-<a href="https://star-history.com/#Wei-Shaw/sub2api&Date">
+<a href="https://star-history.dera.page/#Wei-Shaw/sub2api&Date">
  <picture>
-   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=Wei-Shaw/sub2api&type=Date&theme=dark" />
-   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=Wei-Shaw/sub2api&type=Date" />
-   <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=Wei-Shaw/sub2api&type=Date" />
+   <source media="(prefers-color-scheme: dark)" srcset="https://star-history.dera.page/svg?repos=Wei-Shaw/sub2api&type=Date&theme=dark" />
+   <source media="(prefers-color-scheme: light)" srcset="https://star-history.dera.page/svg?repos=Wei-Shaw/sub2api&type=Date" />
+   <img alt="Star History Chart" src="https://star-history.dera.page/svg?repos=Wei-Shaw/sub2api&type=Date" />
  </picture>
 </a>
 

README_JA.md

@@ -4,7 +4,7 @@
 
 # Sub2API
 
-[![Go](https://img.shields.io/badge/Go-1.26.5-00ADD8.svg)](https://golang.org/)
+[![Go](https://img.shields.io/badge/Go-1.27.0-00ADD8.svg)](https://golang.org/)
 [![Vue](https://img.shields.io/badge/Vue-3.4+-4FC08D.svg)](https://vuejs.org/)
 [![PostgreSQL](https://img.shields.io/badge/PostgreSQL-15+-336791.svg)](https://www.postgresql.org/)
 [![Redis](https://img.shields.io/badge/Redis-7+-DC382D.svg)](https://redis.io/)
@@ -58,6 +58,11 @@
 <td>AIGoCode のご支援に感謝します！AIGoCode は Claude Code、Codex、最新の Gemini モデルを統合したオールインワンプラットフォームで、安定的かつ効率的でコストパフォーマンスに優れた AI コーディングサービスを提供します。柔軟なサブスクリプションプラン、アカウント停止リスクゼロ、VPN 不要の直接アクセス、超高速レスポンスが特長です。AIGoCode は sub2api ユーザー向けに特別特典を用意しています：<a href="https://aigocode.com/invite/SUB2API">こちらのリンク</a>から登録すると、初回チャージ時に 10% のボーナスクレジットを追加プレゼント！</td>
 </tr>
 
+<tr>
+<td width="180"><a href="https://codex-everywhere.com"><img src="assets/partners/logos/codex-everywhere.jpg" alt="CodexEverywhere" width="150"></a></td>
+<td>OpenAI 公式価格のわずか 3% で本物の GPT-5.6 シリーズを提供 — <a href="https://codex-everywhere.com">CodexEverywhere</a> は世界中の開発者にフロンティアモデルへのアクセスを民主化しています。私たちは透明性と誠実さを信条とし、モデル品質は数か月にわたるアクティブなコミュニティの監視によって検証されています。USD および暗号通貨に対応。<a href="https://codex-everywhere.com">codex-everywhere.com</a> で $20 の無料トライアルから始めましょう。</td>
+</tr>
+
 <tr>
 <td width="180"><a href="https://shop.bmoplus.com/?utm_source=github"><img src="assets/partners/logos/bmoplus.jpg" alt="bmoplus" width="150"></a></td>
 <td>本プロジェクトにご支援いただいた BmoPlus に感謝いたします！BmoPlusは、AIサブスクリプションのヘビーユーザー向けに特化した信頼性の高いAIアカウントサービスプロバイダーであり、安定した ChatGPT Plus / ChatGPT Pro (完全保証) / Claude Pro / Super Grok / Gemini Pro の公式代行チャージおよび即納アカウントを提供しています。こちらの<a href="https://shop.bmoplus.com/?utm_source=github">BmoPlus AIアカウント専門店/代行チャージ</a>経由でご登録・ご注文いただいたユーザー様は、GPTを 公式サイト価格の約1割（90% OFF） という驚異的な価格でご利用いただけます！</td>
@@ -119,18 +124,6 @@
 </td>
 </tr>
 
-<tr>
-<td width="180"><a href="https://console.claudeapi.com/agent/register/drTKjyn6wGLK061Z"><img src="assets/partners/logos/claudeapi.jpg" alt="claudeapi" width="150"></a></td>
-<td>Claude API のご支援に感謝します！<a href="https://console.claudeapi.com/agent/register/drTKjyn6wGLK061Z">Claude API</a> は Claude モデルに特化した公式チャネルの API サービスプロバイダーで、Anthropic 公式キーと AWS Bedrock 公式チャネルをベースに、Claude Code や Agent アプリケーションへの安定した接続体験を提供します。Claude 全シリーズのモデルに対応し、Tool Use や長文コンテキストなどの公式機能もそのまま利用可能。リバースエンジニアリングやモデル劣化のないサービスで、Claude Code のヘビーユーザー、Agent エンジニア、企業の技術チームに最適です。<a href="https://console.claudeapi.com/agent/register/drTKjyn6wGLK061Z">専用リンク</a>から登録後カスタマーサポートへご連絡いただくと、無料お試しクレジットを受け取れます。請求書発行やチーム導入にも対応しています。
-</td>
-</tr>
-
-<tr>
-<td width="180"><a href="https://code0.ai/agent/register/LgpIgl9JHtVG53V1?utm_source=zcf&utm_medium=partner&utm_campaign=zcf_2026&utm_content=default"><img src="assets/partners/logos/code0.jpg" alt="code0" width="150"></a></td>
-<td>code0.ai のご支援に感謝します！<a href="https://code0.ai/agent/register/LgpIgl9JHtVG53V1?utm_source=zcf&utm_medium=partner&utm_campaign=zcf_2026&utm_content=default">code0.ai</a> は開発者と技術チーム向けの AI プログラミングワークベンチで、Claude Code や Codex などの主要な Agent コーディング能力を集約し、コード生成、プロジェクト理解、デバッグと修正、コードレビュー、ドキュメント生成といった一般的な開発シーンをサポートします。個人開発者、Agent エンジニア、OSS メンテナー、企業の開発チームに最適で、請求書発行やチーム導入にも対応。<a href="https://code0.ai/agent/register/LgpIgl9JHtVG53V1?utm_source=zcf&utm_medium=partner&utm_campaign=zcf_2026&utm_content=default">専用リンク</a>から登録後カスタマーサポートへご連絡いただくと、無料お試しクレジットを受け取り、より効率的な AI プログラミングワークフローを体験できます。
-</td>
-</tr>
-
 <tr>
 <td width="180"><a href="https://nagora.ai/"><img src="assets/partners/logos/nagora.png" alt="Nagora" width="150"></a></td>
 <td><a href="https://nagora.ai/">Nagora</a>は、開発者やチーム向けに設計されたマルチモデルAI APIゲートウェイです。1つのアカウントと1つのAPIキーだけで、26種類以上の主要なテキストモデルおよび画像モデルを一元的に利用できます。OpenAI、Anthropic、Geminiの各プロトコルに対応し、Claude Code、Codex、Gemini CLIなどの開発ツールにもシームレスに接続できます。 プラットフォームには、インテリジェントルーティング、自動フェイルオーバー、透明性の高い料金体系、請求の一元管理に加え、予算管理、レート制限、同時実行数の制御機能が備わっています。これにより、個人開発、チームでの共同作業、本番環境におけるAI APIの利用を、より安定的かつ柔軟に管理できます。 既存のアプリケーションを改修する必要はありません。Base URLとAPIキーを置き換えるだけで、最短1分で導入を完了できます。</td>
@@ -176,6 +169,11 @@
 <td><a href="https://www.duckip.cn/?keyword=cu7oog6y">DuckIP</a> - 195 以上の国と地域にわたる 9,000 万以上のグローバルレジデンシャルネットワークリソース。ローテーションとスティッキーセッションに対応し、パブリックデータ収集、RAG 更新、モデル評価、マルチリージョンデータワークロードに最適。🟢レジデンシャルプロキシ - 20% オフ；🟢スタティックレジデンシャルプロキシ - ¥50.00/IP から；🟢無制限レジデンシャルプロキシ - ¥19.8/時間 から。✅500M 無料トライアルを取得。</td>
 </tr>
 
+<tr>
+<td width="180"><a href="https://go.apimart.ai/gh-sub2api"><img src="assets/partners/logos/apimart.jpg" alt="APIMart" width="150"></a></td>
+<td>APIMart のご支援に感謝します！<a href="https://go.apimart.ai/gh-sub2api">APIMart</a> は AI 画像・動画生成に特化した低価格 API プラットフォームです。GPT-Image-2 は 1 枚 $0.006 から、1 ドルで 160 枚以上の画像を生成できます。画像と動画の両方に対応する非同期 API を 1 つで利用でき、タスクを送信して ID を取得し、ポーリングまたはコールバックで結果を取得できます。数万枚規模のバッチ処理でもタイムアウトせず、モデルを変更してもコードの変更は不要です。月額料金なしの従量課金制で、<a href="https://go.apimart.ai/gh-sub2api">こちらの登録リンク</a>からすぐに利用を開始できます。</td>
+</tr>
+
 </table>
 
 ## 概要
@@ -207,7 +205,7 @@ Sub2API を拡張・統合するコミュニティプロジェクト:
 
 | コンポーネント | 技術 |
 |-----------|------------|
-| バックエンド | Go 1.26.5, Gin, Ent |
+| バックエンド | Go 1.27.0, Gin, Ent |
 | フロントエンド | Vue 3.4+, Vite 5+, TailwindCSS |
 | データベース | PostgreSQL 15+ |
 | キャッシュ/キュー | Redis 7+ |
@@ -550,12 +548,6 @@ default:
   rate_multiplier: 1.0
 ```
 
-### Sora ステータス（一時的に利用不可）
-
-> ⚠️ Sora 関連の機能は、上流統合およびメディア配信の技術的問題により一時的に利用できません。
-> 現時点では本番環境で Sora に依存しないでください。
-> 既存の `gateway.sora_*` 設定キーは予約されていますが、これらの問題が解決されるまで有効にならない場合があります。
-
 `config.yaml` では追加のセキュリティ関連オプションも利用できます:
 
 - `cors.allowed_origins` - CORS 許可リスト
@@ -732,11 +724,11 @@ sub2api/
 
 ## スター履歴
 
-<a href="https://star-history.com/#Wei-Shaw/sub2api&Date">
+<a href="https://star-history.dera.page/#Wei-Shaw/sub2api&Date">
  <picture>
-   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=Wei-Shaw/sub2api&type=Date&theme=dark" />
-   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=Wei-Shaw/sub2api&type=Date" />
-   <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=Wei-Shaw/sub2api&type=Date" />
+   <source media="(prefers-color-scheme: dark)" srcset="https://star-history.dera.page/svg?repos=Wei-Shaw/sub2api&type=Date&theme=dark" />
+   <source media="(prefers-color-scheme: light)" srcset="https://star-history.dera.page/svg?repos=Wei-Shaw/sub2api&type=Date" />
+   <img alt="Star History Chart" src="https://star-history.dera.page/svg?repos=Wei-Shaw/sub2api&type=Date" />
  </picture>
 </a>
 

backend/.golangci.yml

@@ -121,6 +121,16 @@ linters:
       # Default: true — must be true, ent generates 130K+ lines of code
       generated-is-used: true
 
+  exclusions:
+    rules:
+      # G703/G704 污点分析在测试文件中只会命中 httptest mock 与测试者自设的
+      # 环境变量路径，不构成攻击面；且该分析跨环境结果不稳定（本地/CI 报告的
+      # 位置集合不同），逐点 nolint 无法收敛，故按路径豁免。生产代码不豁免，
+      # 必须逐点 //nolint:gosec 并写明信任边界。
+      - path: '_test\.go$'
+        linters: [ gosec ]
+        text: 'G70[34]'
+
 formatters:
   enable:
     - gofmt

backend/Dockerfile

@@ -1,4 +1,4 @@
-FROM golang:1.26.6-alpine
+FROM golang:1.27.0-alpine
 
 WORKDIR /app
 

backend/cmd/server/VERSION

@@ -1 +1 @@
-0.1.177
+0.1.184

backend/cmd/server/main.go

@@ -153,6 +153,11 @@ func runMainServer() {
 		log.Fatalf("Failed to initialize application: %v", err)
 	}
 	defer app.Cleanup()
+	if app.PluginManager != nil {
+		if err := app.PluginManager.Start(context.Background()); err != nil {
+			log.Printf("Plugin manager started in degraded state: %v", err)
+		}
+	}
 	if app.PromptAudit != nil {
 		if err := app.PromptAudit.Start(context.Background()); err != nil {
 			// Startup continues so unrelated APIs stay up. Fail-closed (unavailable)

backend/cmd/server/wire.go

@@ -25,9 +25,10 @@ import (
 )
 
 type Application struct {
-	Server      *http.Server
-	PromptAudit *securityaudit.PromptService
-	Cleanup     func()
+	Server        *http.Server
+	PromptAudit   *securityaudit.PromptService
+	PluginManager *service.PluginManager
+	Cleanup       func()
 }
 
 func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
@@ -51,12 +52,13 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
 
 		// BuildInfo provider
 		provideServiceBuildInfo,
+		providePluginHostInfo,
 
 		// Cleanup function provider
 		provideCleanup,
 
 		// Application struct
-		wire.Struct(new(Application), "Server", "PromptAudit", "Cleanup"),
+		wire.Struct(new(Application), "Server", "PromptAudit", "PluginManager", "Cleanup"),
 	)
 	return nil, nil
 }
@@ -72,6 +74,13 @@ func provideServiceBuildInfo(buildInfo handler.BuildInfo) service.BuildInfo {
 	}
 }
 
+func providePluginHostInfo(buildInfo handler.BuildInfo) service.PluginHostInfo {
+	return service.PluginHostInfo{
+		Version:   buildInfo.Version,
+		BuildType: buildInfo.BuildType,
+	}
+}
+
 func provideCleanup(
 	entClient *ent.Client,
 	rdb *redis.Client,
@@ -116,7 +125,9 @@ func provideCleanup(
 	upstreamBillingProbe *service.UpstreamBillingProbeService,
 	ollamaCloudUsage *service.OllamaCloudUsageService,
 	auditLog *service.AuditLogService,
+	openAIAutoReset *service.OpenAIQuotaAutoResetService,
 	promptAudit *securityaudit.PromptService,
+	pluginManager *service.PluginManager,
 ) func() {
 	return func() {
 		ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
@@ -129,6 +140,18 @@ func provideCleanup(
 
 		// 应用层清理步骤可并行执行，基础设施资源（Redis/Ent）最后按顺序关闭。
 		parallelSteps := []cleanupStep{
+			{"PluginManager", func() error {
+				if pluginManager != nil {
+					pluginManager.Stop()
+				}
+				return nil
+			}},
+			{"OpenAIQuotaAutoResetService", func() error {
+				if openAIAutoReset != nil {
+					openAIAutoReset.Stop()
+				}
+				return nil
+			}},
 			{"OpsIngressRejectAggregator", func() error {
 				if opsIngressReject != nil {
 					opsIngressReject.Stop()
@@ -328,12 +351,12 @@ func provideCleanup(
 				return nil
 			}},
 			{"ChannelMonitorV2Aggregator", func() error {
-			if channelMonitorV2Aggregator != nil {
-				channelMonitorV2Aggregator.Stop()
-			}
-			return nil
-		}},
-		{"ChannelMonitorRunner", func() error {
+				if channelMonitorV2Aggregator != nil {
+					channelMonitorV2Aggregator.Stop()
+				}
+				return nil
+			}},
+			{"ChannelMonitorRunner", func() error {
 				if channelMonitorRunner != nil {
 					channelMonitorRunner.Stop()
 				}

backend/cmd/server/wire_gen.go

@@ -178,7 +178,7 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
 	channelMonitorUserHandler := handler.NewChannelMonitorUserHandler(channelMonitorService, settingService)
 	channelMonitorV2Repository := repository.NewChannelMonitorV2Repository(db)
 	channelMonitorV2Service := service.ProvideChannelMonitorV2Service(channelMonitorV2Repository, settingService)
-	channelMonitorV2Handler := handler.NewChannelMonitorV2Handler(channelMonitorV2Service)
+	channelMonitorV2Handler := handler.NewChannelMonitorV2Handler(channelMonitorV2Service, apiKeyService)
 	dashboardAggregationRepository := repository.NewDashboardAggregationRepository(db)
 	dashboardStatsCache := repository.NewDashboardCache(redisClient, configConfig)
 	dashboardService := service.NewDashboardService(usageLogRepository, dashboardAggregationRepository, dashboardStatsCache, configConfig)
@@ -194,13 +194,16 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
 	groupCapacityService := service.NewGroupCapacityService(accountRepository, groupRepository, concurrencyService, sessionLimitCache, rpmCache)
 	groupHandler := admin.NewGroupHandler(adminService, dashboardService, groupCapacityService)
 	claudeUsageFetcher := repository.NewClaudeUsageFetcher(httpUpstream)
-	antigravityQuotaFetcher := service.NewAntigravityQuotaFetcher(proxyRepository)
+	antigravityQuotaFetcher := service.NewAntigravityQuotaFetcher(proxyRepository, configConfig)
 	grokQuotaFetcher := service.NewGrokQuotaFetcher()
 	grokQuotaService := service.ProvideGrokQuotaService(accountRepository, proxyRepository, grokTokenProvider, httpUpstream, configConfig, usageLogRepository, settingService)
 	openAIQuotaService := service.ProvideOpenAIQuotaService(accountRepository, proxyRepository, openAITokenProvider, privacyClientFactory, openAIGatewayService)
 	usageCache := service.NewUsageCache()
 	accountUsageService := service.ProvideAccountUsageService(accountRepository, usageLogRepository, claudeUsageFetcher, geminiQuotaService, antigravityQuotaFetcher, grokQuotaFetcher, grokQuotaService, openAIQuotaService, usageCache, identityCache, tlsFingerprintProfileService, openAIGatewayService)
-	accountTestService := service.ProvideAccountTestService(accountRepository, geminiTokenProvider, claudeTokenProvider, grokTokenProvider, antigravityGatewayService, httpUpstream, configConfig, tlsFingerprintProfileService, openAIGatewayService, settingService)
+	pluginRepository := repository.NewPluginRepository(db)
+	pluginHostInfo := providePluginHostInfo(buildInfo)
+	pluginManager := service.NewPluginManager(pluginRepository, secretEncryptor, configConfig, pluginHostInfo)
+	accountTestService := service.ProvideAccountTestService(accountRepository, geminiTokenProvider, claudeTokenProvider, grokTokenProvider, antigravityGatewayService, httpUpstream, configConfig, tlsFingerprintProfileService, openAIGatewayService, settingService, pluginManager)
 	crsSyncService := service.NewCRSSyncService(accountRepository, proxyRepository, oAuthService, openAIOAuthService, geminiOAuthService, configConfig)
 	accountHandler := admin.ProvideAccountHandler(adminService, oAuthService, openAIOAuthService, geminiOAuthService, antigravityOAuthService, grokOAuthService, rateLimitService, accountUsageService, accountTestService, concurrencyService, crsSyncService, sessionLimitCache, rpmCache, compositeTokenCacheInvalidator, grokQuotaService)
 	adminAnnouncementHandler := admin.NewAnnouncementHandler(announcementService)
@@ -251,6 +254,7 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
 	errorPassthroughService := service.NewErrorPassthroughService(errorPassthroughRepository, errorPassthroughCache)
 	errorPassthroughHandler := admin.NewErrorPassthroughHandler(errorPassthroughService)
 	tlsFingerprintProfileHandler := admin.NewTLSFingerprintProfileHandler(tlsFingerprintProfileService)
+	pluginHandler := admin.NewPluginHandler(pluginManager)
 	adminAPIKeyHandler := admin.NewAdminAPIKeyHandler(adminService)
 	scheduledTestPlanRepository := repository.NewScheduledTestPlanRepository(db)
 	scheduledTestResultRepository := repository.NewScheduledTestResultRepository(db)
@@ -280,14 +284,14 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
 	auditLogHandler := admin.NewAuditLogHandler(auditLogService, totpService)
 	upstreamBillingProbeService := service.ProvideUpstreamBillingProbeService(accountRepository, accountTestService, settingService, leaderLockCache, db)
 	ollamaCloudUsageService := service.ProvideOllamaCloudUsageService(accountRepository, httpUpstream, settingService, secretEncryptor, configConfig, leaderLockCache, db)
-	adminHandlers := handler.ProvideAdminHandlers(dashboardHandler, adminUserHandler, groupHandler, accountHandler, adminAnnouncementHandler, dataManagementHandler, backupHandler, oAuthHandler, openAIOAuthHandler, geminiOAuthHandler, antigravityOAuthHandler, grokOAuthHandler, cnProviderHandler, proxyHandler, adminRedeemHandler, promoHandler, settingHandler, opsHandler, systemHandler, adminSubscriptionHandler, adminUsageHandler, userAttributeHandler, errorPassthroughHandler, tlsFingerprintProfileHandler, adminAPIKeyHandler, scheduledTestHandler, channelHandler, channelMonitorHandler, channelMonitorRequestTemplateHandler, contentModerationHandler, promptAdminHandler, paymentHandler, affiliateHandler, complianceHandler, auditLogHandler, upstreamBillingProbeService, ollamaCloudUsageService)
+	adminHandlers := handler.ProvideAdminHandlers(dashboardHandler, adminUserHandler, groupHandler, accountHandler, adminAnnouncementHandler, dataManagementHandler, backupHandler, oAuthHandler, openAIOAuthHandler, geminiOAuthHandler, antigravityOAuthHandler, grokOAuthHandler, cnProviderHandler, proxyHandler, adminRedeemHandler, promoHandler, settingHandler, opsHandler, systemHandler, adminSubscriptionHandler, adminUsageHandler, userAttributeHandler, errorPassthroughHandler, tlsFingerprintProfileHandler, pluginHandler, adminAPIKeyHandler, scheduledTestHandler, channelHandler, channelMonitorHandler, channelMonitorRequestTemplateHandler, contentModerationHandler, promptAdminHandler, paymentHandler, affiliateHandler, complianceHandler, auditLogHandler, upstreamBillingProbeService, ollamaCloudUsageService)
 	usageRecordWorkerPool := service.NewUsageRecordWorkerPool(configConfig)
 	userMsgQueueCache := repository.NewUserMsgQueueCache(redisClient)
 	userMessageQueueService := service.ProvideUserMessageQueueService(userMsgQueueCache, rpmCache, configConfig)
 	legacyEngine := securityaudit.NewLegacyModerationAdapter(contentModerationService)
 	coordinator := securityaudit.NewCoordinator(legacyEngine, promptService)
 	gatewayHandler := handler.ProvideGatewayHandler(gatewayService, openAIGatewayService, geminiMessagesCompatService, antigravityGatewayService, userService, concurrencyService, billingCacheService, usageService, apiKeyService, usageRecordWorkerPool, errorPassthroughService, contentModerationService, userMessageQueueService, configConfig, settingService, coordinator)
-	openAIGatewayHandler := handler.ProvideOpenAIGatewayHandler(openAIGatewayService, concurrencyService, billingCacheService, apiKeyService, usageRecordWorkerPool, errorPassthroughService, contentModerationService, opsService, grokQuotaService, configConfig, coordinator)
+	openAIGatewayHandler := handler.ProvideOpenAIGatewayHandler(openAIGatewayService, pluginManager, concurrencyService, billingCacheService, apiKeyService, usageRecordWorkerPool, errorPassthroughService, contentModerationService, opsService, grokQuotaService, configConfig, coordinator)
 	handlerSettingHandler := handler.ProvideSettingHandler(settingService, buildInfo, notificationEmailService)
 	totpHandler := handler.NewTotpHandler(totpService)
 	passkeyRepository := repository.NewPasskeyRepository(db)
@@ -300,7 +304,8 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
 	handlerPaymentHandler := handler.NewPaymentHandler(paymentService, paymentConfigService)
 	paymentWebhookHandler := handler.NewPaymentWebhookHandler(paymentService, registry)
 	availableChannelHandler := handler.NewAvailableChannelHandler(channelService, apiKeyService, settingService)
-	modelPlazaHandler := handler.NewModelPlazaHandler(channelService, apiKeyService, settingService)
+	modelPlazaService := service.NewModelPlazaService(channelRepository, groupRepository, pricingService, billingService, modelPricingResolver)
+	modelPlazaHandler := handler.NewModelPlazaHandler(modelPlazaService, apiKeyService, settingService)
 	imageTaskStore := repository.NewImageTaskStore(redisClient)
 	imageTaskService := service.ProvideImageTaskService(imageTaskStore, imageStorageSettingService)
 	asyncImageHandler := handler.NewAsyncImageHandler(imageTaskService, openAIGatewayHandler)
@@ -314,7 +319,8 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
 	batchImageHandler := handler.ProvideBatchImageHandler(batchImagePublicService, batchImageDownloadService, batchImageCleanupService, openAIGatewayHandler)
 	idempotencyCoordinator := service.ProvideIdempotencyCoordinator(idempotencyRepository, configConfig)
 	idempotencyCleanupService := service.ProvideIdempotencyCleanupService(idempotencyRepository, configConfig)
-	handlers := handler.ProvideHandlers(authHandler, userHandler, apiKeyHandler, usageHandler, redeemHandler, subscriptionHandler, announcementHandler, channelMonitorUserHandler, channelMonitorV2Handler, adminHandlers, gatewayHandler, openAIGatewayHandler, handlerSettingHandler, totpHandler, passkeyHandler, handlerPaymentHandler, paymentWebhookHandler, availableChannelHandler, modelPlazaHandler, asyncImageHandler, batchImageHandler, idempotencyCoordinator, idempotencyCleanupService)
+	openAIQuotaAutoResetService := service.ProvideOpenAIQuotaAutoResetService(accountRepository, openAIQuotaService, rateLimitService, idempotencyCoordinator, auditLogService, settingService, leaderLockCache)
+	handlers := handler.ProvideHandlers(authHandler, userHandler, apiKeyHandler, usageHandler, redeemHandler, subscriptionHandler, announcementHandler, channelMonitorUserHandler, channelMonitorV2Handler, adminHandlers, gatewayHandler, openAIGatewayHandler, handlerSettingHandler, totpHandler, passkeyHandler, handlerPaymentHandler, paymentWebhookHandler, availableChannelHandler, modelPlazaHandler, asyncImageHandler, batchImageHandler, idempotencyCoordinator, idempotencyCleanupService, openAIQuotaAutoResetService)
 	jwtAuthMiddleware := middleware.NewJWTAuthMiddleware(authService, userService, settingService, auditLogService)
 	optionalJWTAuthMiddleware := middleware.NewOptionalJWTAuthMiddleware(authService, userService, settingService, auditLogService)
 	adminAuthMiddleware := middleware.NewAdminAuthMiddleware(authService, userService, settingService, auditLogService)
@@ -337,15 +343,16 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
 	batchImageWorkerRuntime := service.ProvideBatchImageWorkerRuntime(batchImageRepository, accountRepository, batchImageQueue, usageBillingRepository, usageLogRepository, batchImageModelPricingResolver, apiKeyAuthCacheInvalidator, configConfig)
 	scheduledTestRunnerService := service.ProvideScheduledTestRunnerService(scheduledTestPlanRepository, scheduledTestService, accountTestService, rateLimitService, configConfig)
 	paymentOrderExpiryService := service.ProvidePaymentOrderExpiryService(paymentService, leaderLockCache, db)
-	channelMonitorQuotaFetcher := service.NewChannelMonitorQuotaFetcher(accountUsageService, cnProviderQuotaService, cnProviderBalanceService, accountRepository)
+	channelMonitorQuotaFetcher := service.NewChannelMonitorQuotaFetcher(accountUsageService, cnProviderQuotaService, cnProviderBalanceService, accountRepository, configConfig)
 	channelMonitorRunner := service.ProvideChannelMonitorRunner(channelMonitorService, settingService, channelMonitorQuotaFetcher)
 	channelMonitorV2Aggregator := service.ProvideChannelMonitorV2Aggregator(channelMonitorV2Repository, db, settingService)
 	userPlatformQuotaUsageFlusher := service.ProvideUserPlatformQuotaUsageFlusher(configConfig, billingCache, serviceUserPlatformQuotaRepository, timingWheelService)
-	v := provideCleanup(client, redisClient, opsMetricsCollector, opsAggregationService, opsAlertEvaluatorService, opsCleanupService, opsScheduledReportService, opsSystemLogSink, opsService, opsIngressRejectAggregator, apiKeyService, authCacheInvalidationWorker, schedulerSnapshotService, tokenRefreshService, accountExpiryService, cnProviderBalanceCheckService, openAICodexVersionSyncService, proxyExpiryService, subscriptionExpiryService, usageCleanupService, idempotencyCleanupService, batchImageCleanupService, batchImageWorkerRuntime, pricingService, emailQueueService, billingCacheService, usageRecordWorkerPool, subscriptionService, oAuthService, openAIOAuthService, geminiOAuthService, antigravityOAuthService, grokOAuthService, openAIGatewayService, scheduledTestRunnerService, backupService, paymentOrderExpiryService, channelMonitorRunner, channelMonitorV2Aggregator, userPlatformQuotaUsageFlusher, upstreamBillingProbeService, ollamaCloudUsageService, auditLogService, promptService)
+	v := provideCleanup(client, redisClient, opsMetricsCollector, opsAggregationService, opsAlertEvaluatorService, opsCleanupService, opsScheduledReportService, opsSystemLogSink, opsService, opsIngressRejectAggregator, apiKeyService, authCacheInvalidationWorker, schedulerSnapshotService, tokenRefreshService, accountExpiryService, cnProviderBalanceCheckService, openAICodexVersionSyncService, proxyExpiryService, subscriptionExpiryService, usageCleanupService, idempotencyCleanupService, batchImageCleanupService, batchImageWorkerRuntime, pricingService, emailQueueService, billingCacheService, usageRecordWorkerPool, subscriptionService, oAuthService, openAIOAuthService, geminiOAuthService, antigravityOAuthService, grokOAuthService, openAIGatewayService, scheduledTestRunnerService, backupService, paymentOrderExpiryService, channelMonitorRunner, channelMonitorV2Aggregator, userPlatformQuotaUsageFlusher, upstreamBillingProbeService, ollamaCloudUsageService, auditLogService, openAIQuotaAutoResetService, promptService, pluginManager)
 	application := &Application{
-		Server:      httpServer,
-		PromptAudit: promptService,
-		Cleanup:     v,
+		Server:        httpServer,
+		PromptAudit:   promptService,
+		PluginManager: pluginManager,
+		Cleanup:       v,
 	}
 	return application, nil
 }
@@ -353,9 +360,10 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
 // wire.go:
 
 type Application struct {
-	Server      *http.Server
-	PromptAudit *securityaudit.PromptService
-	Cleanup     func()
+	Server        *http.Server
+	PromptAudit   *securityaudit.PromptService
+	PluginManager *service.PluginManager
+	Cleanup       func()
 }
 
 func providePrivacyClientFactory() service.PrivacyClientFactory {
@@ -369,6 +377,13 @@ func provideServiceBuildInfo(buildInfo handler.BuildInfo) service.BuildInfo {
 	}
 }
 
+func providePluginHostInfo(buildInfo handler.BuildInfo) service.PluginHostInfo {
+	return service.PluginHostInfo{
+		Version:   buildInfo.Version,
+		BuildType: buildInfo.BuildType,
+	}
+}
+
 func provideCleanup(
 	entClient *ent.Client,
 	rdb *redis.Client,
@@ -413,7 +428,9 @@ func provideCleanup(
 	upstreamBillingProbe *service.UpstreamBillingProbeService,
 	ollamaCloudUsage *service.OllamaCloudUsageService,
 	auditLog *service.AuditLogService,
+	openAIAutoReset *service.OpenAIQuotaAutoResetService,
 	promptAudit *securityaudit.PromptService,
+	pluginManager *service.PluginManager,
 ) func() {
 	return func() {
 		ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
@@ -425,6 +442,18 @@ func provideCleanup(
 		}
 
 		parallelSteps := []cleanupStep{
+			{"PluginManager", func() error {
+				if pluginManager != nil {
+					pluginManager.Stop()
+				}
+				return nil
+			}},
+			{"OpenAIQuotaAutoResetService", func() error {
+				if openAIAutoReset != nil {
+					openAIAutoReset.Stop()
+				}
+				return nil
+			}},
 			{"OpsIngressRejectAggregator", func() error {
 				if opsIngressReject != nil {
 					opsIngressReject.Stop()

backend/cmd/server/wire_gen_test.go

@@ -94,7 +94,9 @@ func TestProvideCleanup_WithMinimalDependencies_NoPanic(t *testing.T) {
 		nil, // upstreamBillingProbe
 		nil, // ollamaCloudUsage
 		nil, // auditLog
+		nil, // openAIAutoReset
 		nil, // promptAudit
+		nil, // pluginManager
 	)
 
 	require.NotPanics(t, func() {

backend/ent/group.go

@@ -4,6 +4,7 @@ package ent
 
 import (
 	"encoding/json"
+	"encoding/json/jsontext"
 	"fmt"
 	"strings"
 	"time"
@@ -100,7 +101,7 @@ type Group struct {
 	// 是否按上下文长度应用模型阶梯价格；默认开启以保持官方/渠道长上下文价
 	LongContextPricingEnabled bool `json:"long_context_pricing_enabled,omitempty"`
 	// 分组逐模型定价；优先级高于渠道和内置定价
-	ModelPricing json.RawMessage `json:"model_pricing,omitempty"`
+	ModelPricing jsontext.Value `json:"model_pricing,omitempty"`
 	// 是否仅允许 Claude Code 客户端
 	ClaudeCodeOnly bool `json:"claude_code_only,omitempty"`
 	// 非 Claude Code 请求降级使用的分组 ID

backend/ent/group_create.go

@@ -4,7 +4,7 @@ package ent
 
 import (
 	"context"
-	"encoding/json"
+	"encoding/json/jsontext"
 	"errors"
 	"fmt"
 	"time"
@@ -575,7 +575,7 @@ func (_c *GroupCreate) SetNillableLongContextPricingEnabled(v *bool) *GroupCreat
 }
 
 // SetModelPricing sets the "model_pricing" field.
-func (_c *GroupCreate) SetModelPricing(v json.RawMessage) *GroupCreate {
+func (_c *GroupCreate) SetModelPricing(v jsontext.Value) *GroupCreate {
 	_c.mutation.SetModelPricing(v)
 	return _c
 }
@@ -2445,7 +2445,7 @@ func (u *GroupUpsert) UpdateLongContextPricingEnabled() *GroupUpsert {
 }
 
 // SetModelPricing sets the "model_pricing" field.
-func (u *GroupUpsert) SetModelPricing(v json.RawMessage) *GroupUpsert {
+func (u *GroupUpsert) SetModelPricing(v jsontext.Value) *GroupUpsert {
 	u.Set(group.FieldModelPricing, v)
 	return u
 }
@@ -3615,7 +3615,7 @@ func (u *GroupUpsertOne) UpdateLongContextPricingEnabled() *GroupUpsertOne {
 }
 
 // SetModelPricing sets the "model_pricing" field.
-func (u *GroupUpsertOne) SetModelPricing(v json.RawMessage) *GroupUpsertOne {
+func (u *GroupUpsertOne) SetModelPricing(v jsontext.Value) *GroupUpsertOne {
 	return u.Update(func(s *GroupUpsert) {
 		s.SetModelPricing(v)
 	})
@@ -5005,7 +5005,7 @@ func (u *GroupUpsertBulk) UpdateLongContextPricingEnabled() *GroupUpsertBulk {
 }
 
 // SetModelPricing sets the "model_pricing" field.
-func (u *GroupUpsertBulk) SetModelPricing(v json.RawMessage) *GroupUpsertBulk {
+func (u *GroupUpsertBulk) SetModelPricing(v jsontext.Value) *GroupUpsertBulk {
 	return u.Update(func(s *GroupUpsert) {
 		s.SetModelPricing(v)
 	})

backend/ent/group_update.go

@@ -4,7 +4,7 @@ package ent
 
 import (
 	"context"
-	"encoding/json"
+	"encoding/json/jsontext"
 	"errors"
 	"fmt"
 	"time"
@@ -803,13 +803,13 @@ func (_u *GroupUpdate) SetNillableLongContextPricingEnabled(v *bool) *GroupUpdat
 }
 
 // SetModelPricing sets the "model_pricing" field.
-func (_u *GroupUpdate) SetModelPricing(v json.RawMessage) *GroupUpdate {
+func (_u *GroupUpdate) SetModelPricing(v jsontext.Value) *GroupUpdate {
 	_u.mutation.SetModelPricing(v)
 	return _u
 }
 
 // AppendModelPricing appends value to the "model_pricing" field.
-func (_u *GroupUpdate) AppendModelPricing(v json.RawMessage) *GroupUpdate {
+func (_u *GroupUpdate) AppendModelPricing(v jsontext.Value) *GroupUpdate {
 	_u.mutation.AppendModelPricing(v)
 	return _u
 }
@@ -2924,13 +2924,13 @@ func (_u *GroupUpdateOne) SetNillableLongContextPricingEnabled(v *bool) *GroupUp
 }
 
 // SetModelPricing sets the "model_pricing" field.
-func (_u *GroupUpdateOne) SetModelPricing(v json.RawMessage) *GroupUpdateOne {
+func (_u *GroupUpdateOne) SetModelPricing(v jsontext.Value) *GroupUpdateOne {
 	_u.mutation.SetModelPricing(v)
 	return _u
 }
 
 // AppendModelPricing appends value to the "model_pricing" field.
-func (_u *GroupUpdateOne) AppendModelPricing(v json.RawMessage) *GroupUpdateOne {
+func (_u *GroupUpdateOne) AppendModelPricing(v jsontext.Value) *GroupUpdateOne {
 	_u.mutation.AppendModelPricing(v)
 	return _u
 }

backend/ent/migrate/schema.go

@@ -1801,6 +1801,7 @@ var (
 		{Name: "signup_source", Type: field.TypeString, Default: "email"},
 		{Name: "last_login_at", Type: field.TypeTime, Nullable: true, SchemaType: map[string]string{"postgres": "timestamptz"}},
 		{Name: "last_active_at", Type: field.TypeTime, Nullable: true, SchemaType: map[string]string{"postgres": "timestamptz"}},
+		{Name: "restrict_public_groups", Type: field.TypeBool, Default: false},
 		{Name: "balance_notify_enabled", Type: field.TypeBool, Default: true},
 		{Name: "balance_notify_threshold_type", Type: field.TypeString, Default: "fixed"},
 		{Name: "balance_notify_threshold", Type: field.TypeFloat64, Nullable: true, SchemaType: map[string]string{"postgres": "decimal(20,8)"}},

backend/ent/mutation.go

@@ -4,7 +4,7 @@ package ent
 
 import (
 	"context"
-	"encoding/json"
+	"encoding/json/jsontext"
 	"errors"
 	"fmt"
 	"sync"
@@ -22142,8 +22142,8 @@ type GroupMutation struct {
 	audio_stt_price_per_hour                *float64
 	addaudio_stt_price_per_hour             *float64
 	long_context_pricing_enabled            *bool
-	model_pricing                           *json.RawMessage
-	appendmodel_pricing                     json.RawMessage
+	model_pricing                           *jsontext.Value
+	appendmodel_pricing                     jsontext.Value
 	claude_code_only                        *bool
 	fallback_group_id                       *int64
 	addfallback_group_id                    *int64
@@ -24404,13 +24404,13 @@ func (m *GroupMutation) ResetLongContextPricingEnabled() {
 }
 
 // SetModelPricing sets the "model_pricing" field.
-func (m *GroupMutation) SetModelPricing(jm json.RawMessage) {
-	m.model_pricing = &jm
+func (m *GroupMutation) SetModelPricing(j jsontext.Value) {
+	m.model_pricing = &j
 	m.appendmodel_pricing = nil
 }
 
 // ModelPricing returns the value of the "model_pricing" field in the mutation.
-func (m *GroupMutation) ModelPricing() (r json.RawMessage, exists bool) {
+func (m *GroupMutation) ModelPricing() (r jsontext.Value, exists bool) {
 	v := m.model_pricing
 	if v == nil {
 		return
@@ -24421,7 +24421,7 @@ func (m *GroupMutation) ModelPricing() (r json.RawMessage, exists bool) {
 // OldModelPricing returns the old "model_pricing" field's value of the Group entity.
 // If the Group object wasn't provided to the builder, the object is fetched from the database.
 // An error is returned if the mutation operation is not UpdateOne, or the database query fails.
-func (m *GroupMutation) OldModelPricing(ctx context.Context) (v json.RawMessage, err error) {
+func (m *GroupMutation) OldModelPricing(ctx context.Context) (v jsontext.Value, err error) {
 	if !m.op.Is(OpUpdateOne) {
 		return v, errors.New("OldModelPricing is only allowed on UpdateOne operations")
 	}
@@ -24435,13 +24435,13 @@ func (m *GroupMutation) OldModelPricing(ctx context.Context) (v json.RawMessage,
 	return oldValue.ModelPricing, nil
 }
 
-// AppendModelPricing adds jm to the "model_pricing" field.
-func (m *GroupMutation) AppendModelPricing(jm json.RawMessage) {
-	m.appendmodel_pricing = append(m.appendmodel_pricing, jm...)
+// AppendModelPricing adds j to the "model_pricing" field.
+func (m *GroupMutation) AppendModelPricing(j jsontext.Value) {
+	m.appendmodel_pricing = append(m.appendmodel_pricing, j...)
 }
 
 // AppendedModelPricing returns the list of values that were appended to the "model_pricing" field in this mutation.
-func (m *GroupMutation) AppendedModelPricing() (json.RawMessage, bool) {
+func (m *GroupMutation) AppendedModelPricing() (jsontext.Value, bool) {
 	if len(m.appendmodel_pricing) == 0 {
 		return nil, false
 	}
@@ -26515,7 +26515,7 @@ func (m *GroupMutation) SetField(name string, value ent.Value) error {
 		m.SetLongContextPricingEnabled(v)
 		return nil
 	case group.FieldModelPricing:
-		v, ok := value.(json.RawMessage)
+		v, ok := value.(jsontext.Value)
 		if !ok {
 			return fmt.Errorf("unexpected type %T for field %s", value, name)
 		}
@@ -43151,8 +43151,8 @@ type UsageCleanupTaskMutation struct {
 	created_at      *time.Time
 	updated_at      *time.Time
 	status          *string
-	filters         *json.RawMessage
-	appendfilters   json.RawMessage
+	filters         *jsontext.Value
+	appendfilters   jsontext.Value
 	created_by      *int64
 	addcreated_by   *int64
 	deleted_rows    *int64
@@ -43376,13 +43376,13 @@ func (m *UsageCleanupTaskMutation) ResetStatus() {
 }
 
 // SetFilters sets the "filters" field.
-func (m *UsageCleanupTaskMutation) SetFilters(jm json.RawMessage) {
-	m.filters = &jm
+func (m *UsageCleanupTaskMutation) SetFilters(j jsontext.Value) {
+	m.filters = &j
 	m.appendfilters = nil
 }
 
 // Filters returns the value of the "filters" field in the mutation.
-func (m *UsageCleanupTaskMutation) Filters() (r json.RawMessage, exists bool) {
+func (m *UsageCleanupTaskMutation) Filters() (r jsontext.Value, exists bool) {
 	v := m.filters
 	if v == nil {
 		return
@@ -43393,7 +43393,7 @@ func (m *UsageCleanupTaskMutation) Filters() (r json.RawMessage, exists bool) {
 // OldFilters returns the old "filters" field's value of the UsageCleanupTask entity.
 // If the UsageCleanupTask object wasn't provided to the builder, the object is fetched from the database.
 // An error is returned if the mutation operation is not UpdateOne, or the database query fails.
-func (m *UsageCleanupTaskMutation) OldFilters(ctx context.Context) (v json.RawMessage, err error) {
+func (m *UsageCleanupTaskMutation) OldFilters(ctx context.Context) (v jsontext.Value, err error) {
 	if !m.op.Is(OpUpdateOne) {
 		return v, errors.New("OldFilters is only allowed on UpdateOne operations")
 	}
@@ -43407,13 +43407,13 @@ func (m *UsageCleanupTaskMutation) OldFilters(ctx context.Context) (v json.RawMe
 	return oldValue.Filters, nil
 }
 
-// AppendFilters adds jm to the "filters" field.
-func (m *UsageCleanupTaskMutation) AppendFilters(jm json.RawMessage) {
-	m.appendfilters = append(m.appendfilters, jm...)
+// AppendFilters adds j to the "filters" field.
+func (m *UsageCleanupTaskMutation) AppendFilters(j jsontext.Value) {
+	m.appendfilters = append(m.appendfilters, j...)
 }
 
 // AppendedFilters returns the list of values that were appended to the "filters" field in this mutation.
-func (m *UsageCleanupTaskMutation) AppendedFilters() (json.RawMessage, bool) {
+func (m *UsageCleanupTaskMutation) AppendedFilters() (jsontext.Value, bool) {
 	if len(m.appendfilters) == 0 {
 		return nil, false
 	}
@@ -43964,7 +43964,7 @@ func (m *UsageCleanupTaskMutation) SetField(name string, value ent.Value) error
 		m.SetStatus(v)
 		return nil
 	case usagecleanuptask.FieldFilters:
-		v, ok := value.(json.RawMessage)
+		v, ok := value.(jsontext.Value)
 		if !ok {
 			return fmt.Errorf("unexpected type %T for field %s", value, name)
 		}
@@ -48422,6 +48422,7 @@ type UserMutation struct {
 	signup_source                 *string
 	last_login_at                 *time.Time
 	last_active_at                *time.Time
+	restrict_public_groups        *bool
 	balance_notify_enabled        *bool
 	balance_notify_threshold_type *string
 	balance_notify_threshold      *float64
@@ -49347,6 +49348,42 @@ func (m *UserMutation) ResetLastActiveAt() {
 	delete(m.clearedFields, user.FieldLastActiveAt)
 }
 
+// SetRestrictPublicGroups sets the "restrict_public_groups" field.
+func (m *UserMutation) SetRestrictPublicGroups(b bool) {
+	m.restrict_public_groups = &b
+}
+
+// RestrictPublicGroups returns the value of the "restrict_public_groups" field in the mutation.
+func (m *UserMutation) RestrictPublicGroups() (r bool, exists bool) {
+	v := m.restrict_public_groups
+	if v == nil {
+		return
+	}
+	return *v, true
+}
+
+// OldRestrictPublicGroups returns the old "restrict_public_groups" field's value of the User entity.
+// If the User object wasn't provided to the builder, the object is fetched from the database.
+// An error is returned if the mutation operation is not UpdateOne, or the database query fails.
+func (m *UserMutation) OldRestrictPublicGroups(ctx context.Context) (v bool, err error) {
+	if !m.op.Is(OpUpdateOne) {
+		return v, errors.New("OldRestrictPublicGroups is only allowed on UpdateOne operations")
+	}
+	if m.id == nil || m.oldValue == nil {
+		return v, errors.New("OldRestrictPublicGroups requires an ID field in the mutation")
+	}
+	oldValue, err := m.oldValue(ctx)
+	if err != nil {
+		return v, fmt.Errorf("querying old value for OldRestrictPublicGroups: %w", err)
+	}
+	return oldValue.RestrictPublicGroups, nil
+}
+
+// ResetRestrictPublicGroups resets all changes to the "restrict_public_groups" field.
+func (m *UserMutation) ResetRestrictPublicGroups() {
+	m.restrict_public_groups = nil
+}
+
 // SetBalanceNotifyEnabled sets the "balance_notify_enabled" field.
 func (m *UserMutation) SetBalanceNotifyEnabled(b bool) {
 	m.balance_notify_enabled = &b
@@ -50373,7 +50410,7 @@ func (m *UserMutation) Type() string {
 // order to get all numeric fields that were incremented/decremented, call
 // AddedFields().
 func (m *UserMutation) Fields() []string {
-	fields := make([]string, 0, 24)
+	fields := make([]string, 0, 25)
 	if m.created_at != nil {
 		fields = append(fields, user.FieldCreatedAt)
 	}
@@ -50428,6 +50465,9 @@ func (m *UserMutation) Fields() []string {
 	if m.last_active_at != nil {
 		fields = append(fields, user.FieldLastActiveAt)
 	}
+	if m.restrict_public_groups != nil {
+		fields = append(fields, user.FieldRestrictPublicGroups)
+	}
 	if m.balance_notify_enabled != nil {
 		fields = append(fields, user.FieldBalanceNotifyEnabled)
 	}
@@ -50490,6 +50530,8 @@ func (m *UserMutation) Field(name string) (ent.Value, bool) {
 		return m.LastLoginAt()
 	case user.FieldLastActiveAt:
 		return m.LastActiveAt()
+	case user.FieldRestrictPublicGroups:
+		return m.RestrictPublicGroups()
 	case user.FieldBalanceNotifyEnabled:
 		return m.BalanceNotifyEnabled()
 	case user.FieldBalanceNotifyThresholdType:
@@ -50547,6 +50589,8 @@ func (m *UserMutation) OldField(ctx context.Context, name string) (ent.Value, er
 		return m.OldLastLoginAt(ctx)
 	case user.FieldLastActiveAt:
 		return m.OldLastActiveAt(ctx)
+	case user.FieldRestrictPublicGroups:
+		return m.OldRestrictPublicGroups(ctx)
 	case user.FieldBalanceNotifyEnabled:
 		return m.OldBalanceNotifyEnabled(ctx)
 	case user.FieldBalanceNotifyThresholdType:
@@ -50694,6 +50738,13 @@ func (m *UserMutation) SetField(name string, value ent.Value) error {
 		}
 		m.SetLastActiveAt(v)
 		return nil
+	case user.FieldRestrictPublicGroups:
+		v, ok := value.(bool)
+		if !ok {
+			return fmt.Errorf("unexpected type %T for field %s", value, name)
+		}
+		m.SetRestrictPublicGroups(v)
+		return nil
 	case user.FieldBalanceNotifyEnabled:
 		v, ok := value.(bool)
 		if !ok {
@@ -50953,6 +51004,9 @@ func (m *UserMutation) ResetField(name string) error {
 	case user.FieldLastActiveAt:
 		m.ResetLastActiveAt()
 		return nil
+	case user.FieldRestrictPublicGroups:
+		m.ResetRestrictPublicGroups()
+		return nil
 	case user.FieldBalanceNotifyEnabled:
 		m.ResetBalanceNotifyEnabled()
 		return nil

backend/ent/runtime/runtime.go

@@ -2217,24 +2217,28 @@ func init() {
 	user.DefaultSignupSource = userDescSignupSource.Default.(string)
 	// user.SignupSourceValidator is a validator for the "signup_source" field. It is called by the builders before save.
 	user.SignupSourceValidator = userDescSignupSource.Validators[0].(func(string) error)
+	// userDescRestrictPublicGroups is the schema descriptor for restrict_public_groups field.
+	userDescRestrictPublicGroups := userFields[15].Descriptor()
+	// user.DefaultRestrictPublicGroups holds the default value on creation for the restrict_public_groups field.
+	user.DefaultRestrictPublicGroups = userDescRestrictPublicGroups.Default.(bool)
 	// userDescBalanceNotifyEnabled is the schema descriptor for balance_notify_enabled field.
-	userDescBalanceNotifyEnabled := userFields[15].Descriptor()
+	userDescBalanceNotifyEnabled := userFields[16].Descriptor()
 	// user.DefaultBalanceNotifyEnabled holds the default value on creation for the balance_notify_enabled field.
 	user.DefaultBalanceNotifyEnabled = userDescBalanceNotifyEnabled.Default.(bool)
 	// userDescBalanceNotifyThresholdType is the schema descriptor for balance_notify_threshold_type field.
-	userDescBalanceNotifyThresholdType := userFields[16].Descriptor()
+	userDescBalanceNotifyThresholdType := userFields[17].Descriptor()
 	// user.DefaultBalanceNotifyThresholdType holds the default value on creation for the balance_notify_threshold_type field.
 	user.DefaultBalanceNotifyThresholdType = userDescBalanceNotifyThresholdType.Default.(string)
 	// userDescBalanceNotifyExtraEmails is the schema descriptor for balance_notify_extra_emails field.
-	userDescBalanceNotifyExtraEmails := userFields[18].Descriptor()
+	userDescBalanceNotifyExtraEmails := userFields[19].Descriptor()
 	// user.DefaultBalanceNotifyExtraEmails holds the default value on creation for the balance_notify_extra_emails field.
 	user.DefaultBalanceNotifyExtraEmails = userDescBalanceNotifyExtraEmails.Default.(string)
 	// userDescTotalRecharged is the schema descriptor for total_recharged field.
-	userDescTotalRecharged := userFields[19].Descriptor()
+	userDescTotalRecharged := userFields[20].Descriptor()
 	// user.DefaultTotalRecharged holds the default value on creation for the total_recharged field.
 	user.DefaultTotalRecharged = userDescTotalRecharged.Default.(float64)
 	// userDescRpmLimit is the schema descriptor for rpm_limit field.
-	userDescRpmLimit := userFields[20].Descriptor()
+	userDescRpmLimit := userFields[21].Descriptor()
 	// user.DefaultRpmLimit holds the default value on creation for the rpm_limit field.
 	user.DefaultRpmLimit = userDescRpmLimit.Default.(int)
 	userallowedgroupFields := schema.UserAllowedGroup{}.Fields()

backend/ent/schema/user.go

@@ -96,6 +96,11 @@ func (User) Fields() []ent.Field {
 			Nillable().
 			SchemaType(map[string]string{dialect.Postgres: "timestamptz"}),
 
+		// 公开分组访问限制：为 false 时用户可绑定任意非专属分组（默认行为），
+		// 为 true 时仅可绑定 user_allowed_groups 中列出的公开分组。
+		field.Bool("restrict_public_groups").
+			Default(false),
+
 		// 余额不足通知
 		field.Bool("balance_notify_enabled").
 			Default(true),

backend/ent/usagecleanuptask.go

@@ -4,6 +4,7 @@ package ent
 
 import (
 	"encoding/json"
+	"encoding/json/jsontext"
 	"fmt"
 	"strings"
 	"time"
@@ -25,7 +26,7 @@ type UsageCleanupTask struct {
 	// Status holds the value of the "status" field.
 	Status string `json:"status,omitempty"`
 	// Filters holds the value of the "filters" field.
-	Filters json.RawMessage `json:"filters,omitempty"`
+	Filters jsontext.Value `json:"filters,omitempty"`
 	// CreatedBy holds the value of the "created_by" field.
 	CreatedBy int64 `json:"created_by,omitempty"`
 	// DeletedRows holds the value of the "deleted_rows" field.

backend/ent/usagecleanuptask_create.go

@@ -4,7 +4,7 @@ package ent
 
 import (
 	"context"
-	"encoding/json"
+	"encoding/json/jsontext"
 	"errors"
 	"fmt"
 	"time"
@@ -58,7 +58,7 @@ func (_c *UsageCleanupTaskCreate) SetStatus(v string) *UsageCleanupTaskCreate {
 }
 
 // SetFilters sets the "filters" field.
-func (_c *UsageCleanupTaskCreate) SetFilters(v json.RawMessage) *UsageCleanupTaskCreate {
+func (_c *UsageCleanupTaskCreate) SetFilters(v jsontext.Value) *UsageCleanupTaskCreate {
 	_c.mutation.SetFilters(v)
 	return _c
 }
@@ -375,7 +375,7 @@ func (u *UsageCleanupTaskUpsert) UpdateStatus() *UsageCleanupTaskUpsert {
 }
 
 // SetFilters sets the "filters" field.
-func (u *UsageCleanupTaskUpsert) SetFilters(v json.RawMessage) *UsageCleanupTaskUpsert {
+func (u *UsageCleanupTaskUpsert) SetFilters(v jsontext.Value) *UsageCleanupTaskUpsert {
 	u.Set(usagecleanuptask.FieldFilters, v)
 	return u
 }
@@ -592,7 +592,7 @@ func (u *UsageCleanupTaskUpsertOne) UpdateStatus() *UsageCleanupTaskUpsertOne {
 }
 
 // SetFilters sets the "filters" field.
-func (u *UsageCleanupTaskUpsertOne) SetFilters(v json.RawMessage) *UsageCleanupTaskUpsertOne {
+func (u *UsageCleanupTaskUpsertOne) SetFilters(v jsontext.Value) *UsageCleanupTaskUpsertOne {
 	return u.Update(func(s *UsageCleanupTaskUpsert) {
 		s.SetFilters(v)
 	})
@@ -999,7 +999,7 @@ func (u *UsageCleanupTaskUpsertBulk) UpdateStatus() *UsageCleanupTaskUpsertBulk
 }
 
 // SetFilters sets the "filters" field.
-func (u *UsageCleanupTaskUpsertBulk) SetFilters(v json.RawMessage) *UsageCleanupTaskUpsertBulk {
+func (u *UsageCleanupTaskUpsertBulk) SetFilters(v jsontext.Value) *UsageCleanupTaskUpsertBulk {
 	return u.Update(func(s *UsageCleanupTaskUpsert) {
 		s.SetFilters(v)
 	})

backend/ent/usagecleanuptask_update.go

@@ -4,7 +4,7 @@ package ent
 
 import (
 	"context"
-	"encoding/json"
+	"encoding/json/jsontext"
 	"errors"
 	"fmt"
 	"time"
@@ -51,13 +51,13 @@ func (_u *UsageCleanupTaskUpdate) SetNillableStatus(v *string) *UsageCleanupTask
 }
 
 // SetFilters sets the "filters" field.
-func (_u *UsageCleanupTaskUpdate) SetFilters(v json.RawMessage) *UsageCleanupTaskUpdate {
+func (_u *UsageCleanupTaskUpdate) SetFilters(v jsontext.Value) *UsageCleanupTaskUpdate {
 	_u.mutation.SetFilters(v)
 	return _u
 }
 
 // AppendFilters appends value to the "filters" field.
-func (_u *UsageCleanupTaskUpdate) AppendFilters(v json.RawMessage) *UsageCleanupTaskUpdate {
+func (_u *UsageCleanupTaskUpdate) AppendFilters(v jsontext.Value) *UsageCleanupTaskUpdate {
 	_u.mutation.AppendFilters(v)
 	return _u
 }
@@ -374,13 +374,13 @@ func (_u *UsageCleanupTaskUpdateOne) SetNillableStatus(v *string) *UsageCleanupT
 }
 
 // SetFilters sets the "filters" field.
-func (_u *UsageCleanupTaskUpdateOne) SetFilters(v json.RawMessage) *UsageCleanupTaskUpdateOne {
+func (_u *UsageCleanupTaskUpdateOne) SetFilters(v jsontext.Value) *UsageCleanupTaskUpdateOne {
 	_u.mutation.SetFilters(v)
 	return _u
 }
 
 // AppendFilters appends value to the "filters" field.
-func (_u *UsageCleanupTaskUpdateOne) AppendFilters(v json.RawMessage) *UsageCleanupTaskUpdateOne {
+func (_u *UsageCleanupTaskUpdateOne) AppendFilters(v jsontext.Value) *UsageCleanupTaskUpdateOne {
 	_u.mutation.AppendFilters(v)
 	return _u
 }

backend/ent/user.go

@@ -53,6 +53,8 @@ type User struct {
 	LastLoginAt *time.Time `json:"last_login_at,omitempty"`
 	// LastActiveAt holds the value of the "last_active_at" field.
 	LastActiveAt *time.Time `json:"last_active_at,omitempty"`
+	// RestrictPublicGroups holds the value of the "restrict_public_groups" field.
+	RestrictPublicGroups bool `json:"restrict_public_groups,omitempty"`
 	// BalanceNotifyEnabled holds the value of the "balance_notify_enabled" field.
 	BalanceNotifyEnabled bool `json:"balance_notify_enabled,omitempty"`
 	// BalanceNotifyThresholdType holds the value of the "balance_notify_threshold_type" field.
@@ -237,7 +239,7 @@ func (*User) scanValues(columns []string) ([]any, error) {
 	values := make([]any, len(columns))
 	for i := range columns {
 		switch columns[i] {
-		case user.FieldTotpEnabled, user.FieldBalanceNotifyEnabled:
+		case user.FieldTotpEnabled, user.FieldRestrictPublicGroups, user.FieldBalanceNotifyEnabled:
 			values[i] = new(sql.NullBool)
 		case user.FieldBalance, user.FieldFrozenBalance, user.FieldBalanceNotifyThreshold, user.FieldTotalRecharged:
 			values[i] = new(sql.NullFloat64)
@@ -381,6 +383,12 @@ func (_m *User) assignValues(columns []string, values []any) error {
 				_m.LastActiveAt = new(time.Time)
 				*_m.LastActiveAt = value.Time
 			}
+		case user.FieldRestrictPublicGroups:
+			if value, ok := values[i].(*sql.NullBool); !ok {
+				return fmt.Errorf("unexpected type %T for field restrict_public_groups", values[i])
+			} else if value.Valid {
+				_m.RestrictPublicGroups = value.Bool
+			}
 		case user.FieldBalanceNotifyEnabled:
 			if value, ok := values[i].(*sql.NullBool); !ok {
 				return fmt.Errorf("unexpected type %T for field balance_notify_enabled", values[i])
@@ -588,6 +596,9 @@ func (_m *User) String() string {
 		builder.WriteString(v.Format(time.ANSIC))
 	}
 	builder.WriteString(", ")
+	builder.WriteString("restrict_public_groups=")
+	builder.WriteString(fmt.Sprintf("%v", _m.RestrictPublicGroups))
+	builder.WriteString(", ")
 	builder.WriteString("balance_notify_enabled=")
 	builder.WriteString(fmt.Sprintf("%v", _m.BalanceNotifyEnabled))
 	builder.WriteString(", ")

backend/ent/user/user.go

@@ -51,6 +51,8 @@ const (
 	FieldLastLoginAt = "last_login_at"
 	// FieldLastActiveAt holds the string denoting the last_active_at field in the database.
 	FieldLastActiveAt = "last_active_at"
+	// FieldRestrictPublicGroups holds the string denoting the restrict_public_groups field in the database.
+	FieldRestrictPublicGroups = "restrict_public_groups"
 	// FieldBalanceNotifyEnabled holds the string denoting the balance_notify_enabled field in the database.
 	FieldBalanceNotifyEnabled = "balance_notify_enabled"
 	// FieldBalanceNotifyThresholdType holds the string denoting the balance_notify_threshold_type field in the database.
@@ -212,6 +214,7 @@ var Columns = []string{
 	FieldSignupSource,
 	FieldLastLoginAt,
 	FieldLastActiveAt,
+	FieldRestrictPublicGroups,
 	FieldBalanceNotifyEnabled,
 	FieldBalanceNotifyThresholdType,
 	FieldBalanceNotifyThreshold,
@@ -280,6 +283,8 @@ var (
 	DefaultSignupSource string
 	// SignupSourceValidator is a validator for the "signup_source" field. It is called by the builders before save.
 	SignupSourceValidator func(string) error
+	// DefaultRestrictPublicGroups holds the default value on creation for the "restrict_public_groups" field.
+	DefaultRestrictPublicGroups bool
 	// DefaultBalanceNotifyEnabled holds the default value on creation for the "balance_notify_enabled" field.
 	DefaultBalanceNotifyEnabled bool
 	// DefaultBalanceNotifyThresholdType holds the default value on creation for the "balance_notify_threshold_type" field.
@@ -390,6 +395,11 @@ func ByLastActiveAt(opts ...sql.OrderTermOption) OrderOption {
 	return sql.OrderByField(FieldLastActiveAt, opts...).ToFunc()
 }
 
+// ByRestrictPublicGroups orders the results by the restrict_public_groups field.
+func ByRestrictPublicGroups(opts ...sql.OrderTermOption) OrderOption {
+	return sql.OrderByField(FieldRestrictPublicGroups, opts...).ToFunc()
+}
+
 // ByBalanceNotifyEnabled orders the results by the balance_notify_enabled field.
 func ByBalanceNotifyEnabled(opts ...sql.OrderTermOption) OrderOption {
 	return sql.OrderByField(FieldBalanceNotifyEnabled, opts...).ToFunc()

backend/ent/user/where.go

@@ -145,6 +145,11 @@ func LastActiveAt(v time.Time) predicate.User {
 	return predicate.User(sql.FieldEQ(FieldLastActiveAt, v))
 }
 
+// RestrictPublicGroups applies equality check predicate on the "restrict_public_groups" field. It's identical to RestrictPublicGroupsEQ.
+func RestrictPublicGroups(v bool) predicate.User {
+	return predicate.User(sql.FieldEQ(FieldRestrictPublicGroups, v))
+}
+
 // BalanceNotifyEnabled applies equality check predicate on the "balance_notify_enabled" field. It's identical to BalanceNotifyEnabledEQ.
 func BalanceNotifyEnabled(v bool) predicate.User {
 	return predicate.User(sql.FieldEQ(FieldBalanceNotifyEnabled, v))
@@ -1115,6 +1120,16 @@ func LastActiveAtNotNil() predicate.User {
 	return predicate.User(sql.FieldNotNull(FieldLastActiveAt))
 }
 
+// RestrictPublicGroupsEQ applies the EQ predicate on the "restrict_public_groups" field.
+func RestrictPublicGroupsEQ(v bool) predicate.User {
+	return predicate.User(sql.FieldEQ(FieldRestrictPublicGroups, v))
+}
+
+// RestrictPublicGroupsNEQ applies the NEQ predicate on the "restrict_public_groups" field.
+func RestrictPublicGroupsNEQ(v bool) predicate.User {
+	return predicate.User(sql.FieldNEQ(FieldRestrictPublicGroups, v))
+}
+
 // BalanceNotifyEnabledEQ applies the EQ predicate on the "balance_notify_enabled" field.
 func BalanceNotifyEnabledEQ(v bool) predicate.User {
 	return predicate.User(sql.FieldEQ(FieldBalanceNotifyEnabled, v))

backend/ent/user_create.go

@@ -270,6 +270,20 @@ func (_c *UserCreate) SetNillableLastActiveAt(v *time.Time) *UserCreate {
 	return _c
 }
 
+// SetRestrictPublicGroups sets the "restrict_public_groups" field.
+func (_c *UserCreate) SetRestrictPublicGroups(v bool) *UserCreate {
+	_c.mutation.SetRestrictPublicGroups(v)
+	return _c
+}
+
+// SetNillableRestrictPublicGroups sets the "restrict_public_groups" field if the given value is not nil.
+func (_c *UserCreate) SetNillableRestrictPublicGroups(v *bool) *UserCreate {
+	if v != nil {
+		_c.SetRestrictPublicGroups(*v)
+	}
+	return _c
+}
+
 // SetBalanceNotifyEnabled sets the "balance_notify_enabled" field.
 func (_c *UserCreate) SetBalanceNotifyEnabled(v bool) *UserCreate {
 	_c.mutation.SetBalanceNotifyEnabled(v)
@@ -636,6 +650,10 @@ func (_c *UserCreate) defaults() error {
 		v := user.DefaultSignupSource
 		_c.mutation.SetSignupSource(v)
 	}
+	if _, ok := _c.mutation.RestrictPublicGroups(); !ok {
+		v := user.DefaultRestrictPublicGroups
+		_c.mutation.SetRestrictPublicGroups(v)
+	}
 	if _, ok := _c.mutation.BalanceNotifyEnabled(); !ok {
 		v := user.DefaultBalanceNotifyEnabled
 		_c.mutation.SetBalanceNotifyEnabled(v)
@@ -730,6 +748,9 @@ func (_c *UserCreate) check() error {
 			return &ValidationError{Name: "signup_source", err: fmt.Errorf(`ent: validator failed for field "User.signup_source": %w`, err)}
 		}
 	}
+	if _, ok := _c.mutation.RestrictPublicGroups(); !ok {
+		return &ValidationError{Name: "restrict_public_groups", err: errors.New(`ent: missing required field "User.restrict_public_groups"`)}
+	}
 	if _, ok := _c.mutation.BalanceNotifyEnabled(); !ok {
 		return &ValidationError{Name: "balance_notify_enabled", err: errors.New(`ent: missing required field "User.balance_notify_enabled"`)}
 	}
@@ -844,6 +865,10 @@ func (_c *UserCreate) createSpec() (*User, *sqlgraph.CreateSpec) {
 		_spec.SetField(user.FieldLastActiveAt, field.TypeTime, value)
 		_node.LastActiveAt = &value
 	}
+	if value, ok := _c.mutation.RestrictPublicGroups(); ok {
+		_spec.SetField(user.FieldRestrictPublicGroups, field.TypeBool, value)
+		_node.RestrictPublicGroups = value
+	}
 	if value, ok := _c.mutation.BalanceNotifyEnabled(); ok {
 		_spec.SetField(user.FieldBalanceNotifyEnabled, field.TypeBool, value)
 		_node.BalanceNotifyEnabled = value
@@ -1384,6 +1409,18 @@ func (u *UserUpsert) ClearLastActiveAt() *UserUpsert {
 	return u
 }
 
+// SetRestrictPublicGroups sets the "restrict_public_groups" field.
+func (u *UserUpsert) SetRestrictPublicGroups(v bool) *UserUpsert {
+	u.Set(user.FieldRestrictPublicGroups, v)
+	return u
+}
+
+// UpdateRestrictPublicGroups sets the "restrict_public_groups" field to the value that was provided on create.
+func (u *UserUpsert) UpdateRestrictPublicGroups() *UserUpsert {
+	u.SetExcluded(user.FieldRestrictPublicGroups)
+	return u
+}
+
 // SetBalanceNotifyEnabled sets the "balance_notify_enabled" field.
 func (u *UserUpsert) SetBalanceNotifyEnabled(v bool) *UserUpsert {
 	u.Set(user.FieldBalanceNotifyEnabled, v)
@@ -1819,6 +1856,20 @@ func (u *UserUpsertOne) ClearLastActiveAt() *UserUpsertOne {
 	})
 }
 
+// SetRestrictPublicGroups sets the "restrict_public_groups" field.
+func (u *UserUpsertOne) SetRestrictPublicGroups(v bool) *UserUpsertOne {
+	return u.Update(func(s *UserUpsert) {
+		s.SetRestrictPublicGroups(v)
+	})
+}
+
+// UpdateRestrictPublicGroups sets the "restrict_public_groups" field to the value that was provided on create.
+func (u *UserUpsertOne) UpdateRestrictPublicGroups() *UserUpsertOne {
+	return u.Update(func(s *UserUpsert) {
+		s.UpdateRestrictPublicGroups()
+	})
+}
+
 // SetBalanceNotifyEnabled sets the "balance_notify_enabled" field.
 func (u *UserUpsertOne) SetBalanceNotifyEnabled(v bool) *UserUpsertOne {
 	return u.Update(func(s *UserUpsert) {
@@ -2436,6 +2487,20 @@ func (u *UserUpsertBulk) ClearLastActiveAt() *UserUpsertBulk {
 	})
 }
 
+// SetRestrictPublicGroups sets the "restrict_public_groups" field.
+func (u *UserUpsertBulk) SetRestrictPublicGroups(v bool) *UserUpsertBulk {
+	return u.Update(func(s *UserUpsert) {
+		s.SetRestrictPublicGroups(v)
+	})
+}
+
+// UpdateRestrictPublicGroups sets the "restrict_public_groups" field to the value that was provided on create.
+func (u *UserUpsertBulk) UpdateRestrictPublicGroups() *UserUpsertBulk {
+	return u.Update(func(s *UserUpsert) {
+		s.UpdateRestrictPublicGroups()
+	})
+}
+
 // SetBalanceNotifyEnabled sets the "balance_notify_enabled" field.
 func (u *UserUpsertBulk) SetBalanceNotifyEnabled(v bool) *UserUpsertBulk {
 	return u.Update(func(s *UserUpsert) {

backend/ent/user_update.go

@@ -321,6 +321,20 @@ func (_u *UserUpdate) ClearLastActiveAt() *UserUpdate {
 	return _u
 }
 
+// SetRestrictPublicGroups sets the "restrict_public_groups" field.
+func (_u *UserUpdate) SetRestrictPublicGroups(v bool) *UserUpdate {
+	_u.mutation.SetRestrictPublicGroups(v)
+	return _u
+}
+
+// SetNillableRestrictPublicGroups sets the "restrict_public_groups" field if the given value is not nil.
+func (_u *UserUpdate) SetNillableRestrictPublicGroups(v *bool) *UserUpdate {
+	if v != nil {
+		_u.SetRestrictPublicGroups(*v)
+	}
+	return _u
+}
+
 // SetBalanceNotifyEnabled sets the "balance_notify_enabled" field.
 func (_u *UserUpdate) SetBalanceNotifyEnabled(v bool) *UserUpdate {
 	_u.mutation.SetBalanceNotifyEnabled(v)
@@ -1069,6 +1083,9 @@ func (_u *UserUpdate) sqlSave(ctx context.Context) (_node int, err error) {
 	if _u.mutation.LastActiveAtCleared() {
 		_spec.ClearField(user.FieldLastActiveAt, field.TypeTime)
 	}
+	if value, ok := _u.mutation.RestrictPublicGroups(); ok {
+		_spec.SetField(user.FieldRestrictPublicGroups, field.TypeBool, value)
+	}
 	if value, ok := _u.mutation.BalanceNotifyEnabled(); ok {
 		_spec.SetField(user.FieldBalanceNotifyEnabled, field.TypeBool, value)
 	}
@@ -1997,6 +2014,20 @@ func (_u *UserUpdateOne) ClearLastActiveAt() *UserUpdateOne {
 	return _u
 }
 
+// SetRestrictPublicGroups sets the "restrict_public_groups" field.
+func (_u *UserUpdateOne) SetRestrictPublicGroups(v bool) *UserUpdateOne {
+	_u.mutation.SetRestrictPublicGroups(v)
+	return _u
+}
+
+// SetNillableRestrictPublicGroups sets the "restrict_public_groups" field if the given value is not nil.
+func (_u *UserUpdateOne) SetNillableRestrictPublicGroups(v *bool) *UserUpdateOne {
+	if v != nil {
+		_u.SetRestrictPublicGroups(*v)
+	}
+	return _u
+}
+
 // SetBalanceNotifyEnabled sets the "balance_notify_enabled" field.
 func (_u *UserUpdateOne) SetBalanceNotifyEnabled(v bool) *UserUpdateOne {
 	_u.mutation.SetBalanceNotifyEnabled(v)
@@ -2775,6 +2806,9 @@ func (_u *UserUpdateOne) sqlSave(ctx context.Context) (_node *User, err error) {
 	if _u.mutation.LastActiveAtCleared() {
 		_spec.ClearField(user.FieldLastActiveAt, field.TypeTime)
 	}
+	if value, ok := _u.mutation.RestrictPublicGroups(); ok {
+		_spec.SetField(user.FieldRestrictPublicGroups, field.TypeBool, value)
+	}
 	if value, ok := _u.mutation.BalanceNotifyEnabled(); ok {
 		_spec.SetField(user.FieldBalanceNotifyEnabled, field.TypeBool, value)
 	}

backend/go.mod

@@ -1,6 +1,6 @@
 module github.com/Wei-Shaw/sub2api
 
-go 1.26.6
+go 1.27.0
 
 require (
 	entgo.io/ent v0.14.5
@@ -24,6 +24,8 @@ require (
 	github.com/google/uuid v1.6.0
 	github.com/google/wire v0.7.0
 	github.com/gorilla/websocket v1.5.3
+	github.com/hashicorp/go-hclog v1.6.3
+	github.com/hashicorp/go-plugin v1.8.0
 	github.com/imroc/req/v3 v3.59.0
 	github.com/klauspost/compress v1.18.2
 	github.com/lib/pq v1.10.9
@@ -54,6 +56,8 @@ require (
 	golang.org/x/net v0.56.0
 	golang.org/x/sync v0.21.0
 	golang.org/x/term v0.44.0
+	google.golang.org/grpc v1.82.1
+	google.golang.org/protobuf v1.36.11
 	gopkg.in/natefinch/lumberjack.v2 v2.2.1
 	gopkg.in/yaml.v3 v3.0.1
 	modernc.org/sqlite v1.44.3
@@ -121,12 +125,14 @@ require (
 	github.com/go-viper/mapstructure/v2 v2.5.0 // indirect
 	github.com/go-webauthn/x v0.2.6 // indirect
 	github.com/goccy/go-json v0.10.2 // indirect
+	github.com/golang/protobuf v1.5.4 // indirect
 	github.com/google/go-cmp v0.7.0 // indirect
 	github.com/google/go-querystring v1.1.0 // indirect
 	github.com/google/go-tpm v0.9.8 // indirect
 	github.com/grpc-ecosystem/grpc-gateway/v2 v2.27.3 // indirect
 	github.com/hashicorp/hcl v1.0.0 // indirect
 	github.com/hashicorp/hcl/v2 v2.18.1 // indirect
+	github.com/hashicorp/yamux v0.1.2 // indirect
 	github.com/icholy/digest v1.1.0 // indirect
 	github.com/json-iterator/go v1.1.12 // indirect
 	github.com/klauspost/cpuid/v2 v2.2.4 // indirect
@@ -149,6 +155,7 @@ require (
 	github.com/modern-go/reflect2 v1.0.2 // indirect
 	github.com/morikuni/aec v1.0.0 // indirect
 	github.com/ncruces/go-strftime v1.0.0 // indirect
+	github.com/oklog/run v1.1.0 // indirect
 	github.com/opencontainers/go-digest v1.0.0 // indirect
 	github.com/opencontainers/image-spec v1.1.1 // indirect
 	github.com/pelletier/go-toml/v2 v2.2.2 // indirect
@@ -187,10 +194,9 @@ require (
 	github.com/zclconf/go-cty-yaml v1.1.0 // indirect
 	go.opentelemetry.io/auto/sdk v1.2.1 // indirect
 	go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.49.0 // indirect
-	go.opentelemetry.io/otel v1.41.0 // indirect
-	go.opentelemetry.io/otel/metric v1.41.0 // indirect
-	go.opentelemetry.io/otel/sdk v1.41.0 // indirect
-	go.opentelemetry.io/otel/trace v1.41.0 // indirect
+	go.opentelemetry.io/otel v1.43.0 // indirect
+	go.opentelemetry.io/otel/metric v1.43.0 // indirect
+	go.opentelemetry.io/otel/trace v1.43.0 // indirect
 	go.uber.org/atomic v1.10.0 // indirect
 	go.uber.org/automaxprocs v1.6.0 // indirect
 	go.uber.org/multierr v1.9.0 // indirect
@@ -200,8 +206,7 @@ require (
 	golang.org/x/text v0.39.0 // indirect
 	golang.org/x/time v0.12.0 // indirect
 	golang.org/x/tools v0.47.0 // indirect
-	google.golang.org/grpc v1.75.1 // indirect
-	google.golang.org/protobuf v1.36.10 // indirect
+	google.golang.org/genproto/googleapis/rpc v0.0.0-20260414002931-afd174a4e478 // indirect
 	gopkg.in/ini.v1 v1.67.0 // indirect
 	modernc.org/libc v1.67.6 // indirect
 	modernc.org/mathutil v1.7.1 // indirect

backend/go.sum

@@ -120,6 +120,8 @@ github.com/bsm/ginkgo/v2 v2.12.0 h1:Ny8MWAHyOepLGlLKYmXG4IEkioBysk6GpaRTLC8zwWs=
 github.com/bsm/ginkgo/v2 v2.12.0/go.mod h1:SwYbGRRDovPVboqFv0tPTcG1sN61LM1Z4ARdbAV9g4c=
 github.com/bsm/gomega v1.27.10 h1:yeMWxP2pV2fG3FgAODIY8EiRE3dy0aeFYt4l7wh6yKA=
 github.com/bsm/gomega v1.27.10/go.mod h1:JyEr/xRbxbtgWNi8tIEVPUYZ5Dzef52k01W3YH0H+O0=
+github.com/bufbuild/protocompile v0.14.1 h1:iA73zAf/fyljNjQKwYzUHD6AD4R8KMasmwa/FBatYVw=
+github.com/bufbuild/protocompile v0.14.1/go.mod h1:ppVdAIhbr2H8asPk6k4pY7t9zB1OU5DoEw9xY/FUi1c=
 github.com/bytedance/sonic v1.5.0/go.mod h1:ED5hyg4y6t3/9Ku1R6dU/4KyJ48DZ4jPhfY1O2AihPM=
 github.com/bytedance/sonic v1.9.1 h1:6iJ6NqdoxCDr6mbY8h18oSO+cShGSMRGCEo7F2h0x8s=
 github.com/bytedance/sonic v1.9.1/go.mod h1:i736AoUSYt75HyZLoJW9ERYxcy6eaN6h4BZXU064P/U=
@@ -176,6 +178,7 @@ github.com/ebitengine/purego v0.8.4/go.mod h1:iIjxzd6CiRiOG0UyXP+V1+jWqUXVjPKLAI
 github.com/envoyproxy/go-control-plane v0.9.0/go.mod h1:YTl/9mNaCwkRvm6d1a2C3ymFceY/DCBVvsKhRF0iEA4=
 github.com/envoyproxy/go-control-plane v0.9.4/go.mod h1:6rpuAdCZL397s3pYoYcLgu1mIlRU8Am5FuJP05cCM98=
 github.com/envoyproxy/protoc-gen-validate v0.1.0/go.mod h1:iSmxcyjqTsJpI2R4NaDN7+kN2VEUnK/pcBlmesArF7c=
+github.com/fatih/color v1.13.0/go.mod h1:kLAiJbzzSOZDVNGyDpeOxJ47H46qBXwg5ILebYFFOfk=
 github.com/fatih/color v1.18.0 h1:S8gINlzdQ840/4pfAwic/ZE0djQEH3wM94VfqLTZcOM=
 github.com/fatih/color v1.18.0/go.mod h1:4FelSpRwEGDpQ12mAdzqdOukCy4u8WUtOY6lkT/6HfU=
 github.com/felixge/httpsnoop v1.0.4 h1:NFTV2Zj1bL4mc9sqWACXbQFVBBg2W3GPvqp8/ESS2Wg=
@@ -232,6 +235,8 @@ github.com/golang/protobuf v1.4.0-rc.2/go.mod h1:LlEzMj4AhA7rCAGe4KMBDvJI+AwstrU
 github.com/golang/protobuf v1.4.0-rc.4.0.20200313231945-b860323f09d0/go.mod h1:WU3c8KckQ9AFe+yFwt9sWVRKCVIyN9cPHBJSNnbL67w=
 github.com/golang/protobuf v1.4.0/go.mod h1:jodUvKwWbYaEsadDk5Fwe5c77LiNKVO9IDvqG2KuDX0=
 github.com/golang/protobuf v1.4.2/go.mod h1:oDoupMAO8OvCJWAcko0GGGIgR6R6ocIYbsSw735rRwI=
+github.com/golang/protobuf v1.5.4 h1:i7eJL8qZTpSEXOPTxNKhASYpMn+8e5Q6AdndVa1dWek=
+github.com/golang/protobuf v1.5.4/go.mod h1:lnTiLA8Wa4RWRcIUkrtSVa5nRhsEGBg48fD6rSs7xps=
 github.com/google/go-cmp v0.2.0/go.mod h1:oXzfMopK8JAjlY9xF4vHSVASa0yLyX7SntLO5aqRK0M=
 github.com/google/go-cmp v0.3.0/go.mod h1:8QqcDgzrUqlUb/G2PQTWiueGozuR1884gddMywk6iLU=
 github.com/google/go-cmp v0.3.1/go.mod h1:8QqcDgzrUqlUb/G2PQTWiueGozuR1884gddMywk6iLU=
@@ -250,8 +255,6 @@ github.com/google/go-tpm-tools v0.3.13-0.20230620182252-4639ecce2aba/go.mod h1:E
 github.com/google/gofuzz v1.0.0/go.mod h1:dBl0BpW6vV/+mYPU4Po3pmUjxk6FQPldtuIdl/M65Eg=
 github.com/google/pprof v0.0.0-20250317173921-a4b03ec1a45e h1:ijClszYn+mADRFY17kjQEVQ1XRhq2/JR1M3sGqeJoxs=
 github.com/google/pprof v0.0.0-20250317173921-a4b03ec1a45e/go.mod h1:boTsfXsheKC2y+lKOCMpSfarhxDeIzfZG1jqGcPl3cA=
-github.com/google/subcommands v1.2.0 h1:vWQspBTo2nEqTUFita5/KeEWlUL8kQObDFbub/EN9oE=
-github.com/google/subcommands v1.2.0/go.mod h1:ZjhPrFU+Olkh9WazFPsl27BQ4UPiG37m3yTrtFlrHVk=
 github.com/google/uuid v1.6.0 h1:NIvaJDMOsjHA8n1jAhLSgzrAzy1Hgr+hNrb57e+94F0=
 github.com/google/uuid v1.6.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
 github.com/google/wire v0.7.0 h1:JxUKI6+CVBgCO2WToKy/nQk0sS+amI9z9EjVmdaocj4=
@@ -262,6 +265,10 @@ github.com/gorilla/websocket v1.5.3 h1:saDtZ6Pbx/0u+bgYQ3q96pZgCzfhKXGPqt7kZ72aN
 github.com/gorilla/websocket v1.5.3/go.mod h1:YR8l580nyteQvAITg2hZ9XVh4b55+EU/adAjf1fMHhE=
 github.com/grpc-ecosystem/grpc-gateway/v2 v2.27.3 h1:NmZ1PKzSTQbuGHw9DGPFomqkkLWMC+vZCkfs+FHv1Vg=
 github.com/grpc-ecosystem/grpc-gateway/v2 v2.27.3/go.mod h1:zQrxl1YP88HQlA6i9c63DSVPFklWpGX4OWAc9bFuaH4=
+github.com/hashicorp/go-hclog v1.6.3 h1:Qr2kF+eVWjTiYmU7Y31tYlP1h0q/X3Nl3tPGdaB11/k=
+github.com/hashicorp/go-hclog v1.6.3/go.mod h1:W4Qnvbt70Wk/zYJryRzDRU/4r0kIg0PVHBcfoyhpF5M=
+github.com/hashicorp/go-plugin v1.8.0 h1:ie8S6RRY8RvB2usYZv+AAZ/wBvx2AU5p5QeP5j/FORs=
+github.com/hashicorp/go-plugin v1.8.0/go.mod h1:BExt6KEaIYx804z8k4gRzRLEvxKVb+kn0NMcihqOqb8=
 github.com/hashicorp/golang-lru v0.5.4 h1:YDjusn29QI/Das2iO9M0BHnIbxPeyuCHsjMW+lJfyTc=
 github.com/hashicorp/golang-lru/v2 v2.0.7 h1:a+bsQ5rvGLjzHuww6tVxozPZFVghXaHOwFs4luLUK2k=
 github.com/hashicorp/golang-lru/v2 v2.0.7/go.mod h1:QeFd9opnmA6QUJc5vARoKUSoFhyfM2/ZepoAG6RGpeM=
@@ -269,6 +276,8 @@ github.com/hashicorp/hcl v1.0.0 h1:0Anlzjpi4vEasTeNFn2mLJgTSwt0+6sfsiTG8qcWGx4=
 github.com/hashicorp/hcl v1.0.0/go.mod h1:E5yfLk+7swimpb2L/Alb/PJmXilQ/rhwaUYs4T20WEQ=
 github.com/hashicorp/hcl/v2 v2.18.1 h1:6nxnOJFku1EuSawSD81fuviYUV8DxFr3fp2dUi3ZYSo=
 github.com/hashicorp/hcl/v2 v2.18.1/go.mod h1:ThLC89FV4p9MPW804KVbe/cEXoQ8NZEh+JtMeeGErHE=
+github.com/hashicorp/yamux v0.1.2 h1:XtB8kyFOyHXYVFnwT5C3+Bdo8gArse7j2AQ0DA0Uey8=
+github.com/hashicorp/yamux v0.1.2/go.mod h1:C+zze2n6e/7wshOZep2A70/aQU6QBRWJO/G6FT1wIns=
 github.com/icholy/digest v1.1.0 h1:HfGg9Irj7i+IX1o1QAmPfIBNu/Q5A5Tu3n/MED9k9H4=
 github.com/icholy/digest v1.1.0/go.mod h1:QNrsSGQ5v7v9cReDI0+eyjsXGUoRSUZQHeQ5C4XLa0Y=
 github.com/imroc/req/v3 v3.59.0 h1:PqKhJHyBmJYob47LVuTHwRZE00ZO6icbLHe5Zra13jo=
@@ -281,6 +290,8 @@ github.com/jackc/pgx/v5 v5.7.4 h1:9wKznZrhWa2QiHL+NjTSPP6yjl3451BX3imWDnokYlg=
 github.com/jackc/pgx/v5 v5.7.4/go.mod h1:ncY89UGWxg82EykZUwSpUKEfccBGGYq1xjrOpsbsfGQ=
 github.com/jackc/puddle/v2 v2.2.2 h1:PR8nw+E/1w0GLuRFSmiioY6UooMp6KJv0/61nB7icHo=
 github.com/jackc/puddle/v2 v2.2.2/go.mod h1:vriiEXHvEE654aYKXXjOvZM39qJ0q+azkZFrfEOc3H4=
+github.com/jhump/protoreflect v1.17.0 h1:qOEr613fac2lOuTgWN4tPAtLL7fUSbuJL5X5XumQh94=
+github.com/jhump/protoreflect v1.17.0/go.mod h1:h9+vUUL38jiBzck8ck+6G/aeMX8Z4QUY/NiJPwPNi+8=
 github.com/json-iterator/go v1.1.10/go.mod h1:KdQUCv79m/52Kvf8AW2vK1V8akMuk1QjK/uOdHXbAo4=
 github.com/json-iterator/go v1.1.12 h1:PV8peI4a0ysnczrg+LtxykD8LfKY9ML6u2jnxaEnrnM=
 github.com/json-iterator/go v1.1.12/go.mod h1:e30LSqwooZae/UwlEbR2852Gd8hjQvJoHmT4TnhNGBo=
@@ -307,13 +318,15 @@ github.com/lufia/plan9stats v0.0.0-20211012122336-39d0f177ccd0 h1:6E+4a0GO5zZEnZ
 github.com/lufia/plan9stats v0.0.0-20211012122336-39d0f177ccd0/go.mod h1:zJYVVT2jmtg6P3p1VtQj7WsuWi/y4VnjVBn7F8KPB3I=
 github.com/magiconair/properties v1.8.10 h1:s31yESBquKXCV9a/ScB3ESkOjUYYv+X0rg8SYxI99mE=
 github.com/magiconair/properties v1.8.10/go.mod h1:Dhd985XPs7jluiymwWYZ0G4Z61jb3vdS329zhj2hYo0=
+github.com/mattn/go-colorable v0.1.9/go.mod h1:u6P/XSegPjTcexA+o6vUJrdnUu04hMope9wVRipJSqc=
+github.com/mattn/go-colorable v0.1.12/go.mod h1:u5H1YNBxpqRaxsYJYSkiCWKzEfiAb1Gb520KVy5xxl4=
 github.com/mattn/go-colorable v0.1.13 h1:fFA4WZxdEF4tXPZVKMLwD8oUnCTTo08duU7wxecdEvA=
 github.com/mattn/go-colorable v0.1.13/go.mod h1:7S9/ev0klgBDR4GtXTXX8a3vIGJpMovkB8vQcUbaXHg=
+github.com/mattn/go-isatty v0.0.12/go.mod h1:cbi8OIDigv2wuxKPP5vlRcQ1OAZbq2CE4Kysco4FUpU=
+github.com/mattn/go-isatty v0.0.14/go.mod h1:7GGIvUiUoEMVVmxf/4nioHXj79iQHKdU27kJ6hsGG94=
 github.com/mattn/go-isatty v0.0.16/go.mod h1:kYGgaQfpe5nmfYZH+SKPsOc2e4SrIfOl2e/yFXSvRLM=
 github.com/mattn/go-isatty v0.0.20 h1:xfD0iDuEKnDkl03q4limB+vH+GxLEtL/jb4xVJSWWEY=
 github.com/mattn/go-isatty v0.0.20/go.mod h1:W+V8PltTTMOvKvAeJH7IuucS94S2C6jfK/D7dTCTo3Y=
-github.com/mattn/go-runewidth v0.0.15 h1:UNAjwbU9l54TA3KzvqLGxwWjHmMgBUVhBiTjelZgg3U=
-github.com/mattn/go-runewidth v0.0.15/go.mod h1:Jdepj2loyihRzMpdS35Xk/zdY8IAYHsh153qUoGf23w=
 github.com/mattn/go-sqlite3 v1.14.17 h1:mCRHCLDUBXgpKAqIKsaAaAsrAlbkeomtRFKXh2L6YIM=
 github.com/mattn/go-sqlite3 v1.14.17/go.mod h1:2eHXhiwb8IkHr+BDWZGa96P6+rkvnG63S2DGjv9HUNg=
 github.com/mdelapenya/tlscert v0.2.0 h1:7H81W6Z/4weDvZBNOfQte5GpIMo0lGYEeWbkGp5LJHI=
@@ -350,8 +363,8 @@ github.com/morikuni/aec v1.0.0/go.mod h1:BbKIizmSmc5MMPqRYbxO4ZU0S0+P200+tUnFx7P
 github.com/ncruces/go-strftime v1.0.0 h1:HMFp8mLCTPp341M/ZnA4qaf7ZlsbTc+miZjCLOFAw7w=
 github.com/ncruces/go-strftime v1.0.0/go.mod h1:Fwc5htZGVVkseilnfgOVb9mKy6w1naJmn9CehxcKcls=
 github.com/niemeyer/pretty v0.0.0-20200227124842-a10e7caefd8e/go.mod h1:zD1mROLANZcx1PVRCS0qkT7pwLkGfwJo4zjcN/Tysno=
-github.com/olekukonko/tablewriter v0.0.5 h1:P2Ga83D34wi1o9J6Wh1mRuqd4mF/x/lgBS7N7AbDhec=
-github.com/olekukonko/tablewriter v0.0.5/go.mod h1:hPp6KlRPjbx+hW8ykQs1w3UBbZlj6HuIJcUGPhkA7kY=
+github.com/oklog/run v1.1.0 h1:GEenZ1cK0+q0+wsJew9qUg/DyD8k3JzYsZAi5gYi2mA=
+github.com/oklog/run v1.1.0/go.mod h1:sVPdnTZT1zYwAJeCMu2Th4T21pA3FPOQRfWjQlk7DVU=
 github.com/opencontainers/go-digest v1.0.0 h1:apOUWs51W5PlhuyGyz9FCeeBIOUDA/6nW8Oi/yOhh5U=
 github.com/opencontainers/go-digest v1.0.0/go.mod h1:0JzlMkj0TRzQZfJkVvzbP0HBR3IKzErnv2BNG4W4MAM=
 github.com/opencontainers/image-spec v1.1.1 h1:y0fUlFfIZhPF1W537XOLg0/fcx6zcHCJwooC2xJA040=
@@ -386,8 +399,6 @@ github.com/refraction-networking/utls v1.8.2 h1:j4Q1gJj0xngdeH+Ox/qND11aEfhpgoEv
 github.com/refraction-networking/utls v1.8.2/go.mod h1:jkSOEkLqn+S/jtpEHPOsVv/4V4EVnelwbMQl4vCWXAM=
 github.com/remyoudompheng/bigfft v0.0.0-20230129092748-24d4a6f8daec h1:W09IVJc94icq4NjY3clb7Lk8O1qJ8BdBEF8z0ibU0rE=
 github.com/remyoudompheng/bigfft v0.0.0-20230129092748-24d4a6f8daec/go.mod h1:qqbHyh8v60DhA7CoWK5oRCqLrMHRGoxYCSS9EjAz6Eo=
-github.com/rivo/uniseg v0.2.0 h1:S1pD9weZBuJdFmowNwbpi7BJ8TNftyUImj/0WQi72jY=
-github.com/rivo/uniseg v0.2.0/go.mod h1:J6wj4VEh+S6ZtnVlnTBMWIodfgj8LQOQFoIToxlJtxc=
 github.com/robfig/cron/v3 v3.0.1 h1:WdRxkvbJztn8LMz/QEvLN5sBU+xKpSqwwUO1Pjr4qDs=
 github.com/robfig/cron/v3 v3.0.1/go.mod h1:eQICP3HwyT7UooqI/z+Ov+PtYAWygg1TEWWzGIFLtro=
 github.com/rogpeppe/go-internal v1.14.1 h1:UQB4HGPB6osV0SQTLymcB4TgvyWu6ZyliaW0tI/otEQ=
@@ -423,8 +434,6 @@ github.com/spf13/afero v1.11.0 h1:WJQKhtpdm3v2IzqG8VMqrr6Rf3UYpEF239Jy9wNepM8=
 github.com/spf13/afero v1.11.0/go.mod h1:GH9Y3pIexgf1MTIWtNGyogA5MwRIDXGUr+hbWNoBjkY=
 github.com/spf13/cast v1.6.0 h1:GEiTHELF+vaR5dhz3VqZfFSzZjYbgeKDpBxQVS4GYJ0=
 github.com/spf13/cast v1.6.0/go.mod h1:ancEpBxwJDODSW/UG4rDrAqiKolqNNh2DX3mk86cAdo=
-github.com/spf13/cobra v1.7.0 h1:hyqWnYt1ZQShIddO5kBpj3vu05/++x6tJ6dg8EC572I=
-github.com/spf13/cobra v1.7.0/go.mod h1:uLxZILRyS/50WlhOIKD7W6V5bgeIt+4sICxh6uRMrb0=
 github.com/spf13/pflag v1.0.5 h1:iy+VFUOCP1a+8yFto/drg2CJ5u0yRoB7fZw3DKv/JXA=
 github.com/spf13/pflag v1.0.5/go.mod h1:McXfInJRrz4CZXVZOBLb0bTZqETkiAhM9Iw0y3An2Bg=
 github.com/spf13/viper v1.18.2 h1:LUXCnvUvSM6FXAsj6nnfc8Q2tp1dIgUfY9Kc8GsSOiQ=
@@ -439,6 +448,7 @@ github.com/stretchr/testify v1.3.0/go.mod h1:M5WIy9Dh21IEIfnGCwXGc5bZfKNJtfHm1UV
 github.com/stretchr/testify v1.5.1/go.mod h1:5W2xD1RspED5o8YsWQXVCued0rvSQ+mT+I5cxcmMvtA=
 github.com/stretchr/testify v1.7.0/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
 github.com/stretchr/testify v1.7.1/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
+github.com/stretchr/testify v1.7.2/go.mod h1:R6va5+xMeoiuVRoj+gSkQ7d3FALtqAAGI1FQKckRals=
 github.com/stretchr/testify v1.8.0/go.mod h1:yNjHg4UonilssWZ8iaSj1OCr/vHnekPRkoO+kdMU+MU=
 github.com/stretchr/testify v1.8.1/go.mod h1:w2LPCIKwWwSfY2zedu0+kehJoqGctiVI29o6fzry7u4=
 github.com/stretchr/testify v1.8.2/go.mod h1:w2LPCIKwWwSfY2zedu0+kehJoqGctiVI29o6fzry7u4=
@@ -507,18 +517,20 @@ go.opentelemetry.io/auto/sdk v1.2.1 h1:jXsnJ4Lmnqd11kwkBV2LgLoFMZKizbCi5fNZ/ipaZ
 go.opentelemetry.io/auto/sdk v1.2.1/go.mod h1:KRTj+aOaElaLi+wW1kO/DZRXwkF4C5xPbEe3ZiIhN7Y=
 go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.49.0 h1:jq9TW8u3so/bN+JPT166wjOI6/vQPF6Xe7nMNIltagk=
 go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.49.0/go.mod h1:p8pYQP+m5XfbZm9fxtSKAbM6oIllS7s2AfxrChvc7iw=
-go.opentelemetry.io/otel v1.41.0 h1:YlEwVsGAlCvczDILpUXpIpPSL/VPugt7zHThEMLce1c=
-go.opentelemetry.io/otel v1.41.0/go.mod h1:Yt4UwgEKeT05QbLwbyHXEwhnjxNO6D8L5PQP51/46dE=
+go.opentelemetry.io/otel v1.43.0 h1:mYIM03dnh5zfN7HautFE4ieIig9amkNANT+xcVxAj9I=
+go.opentelemetry.io/otel v1.43.0/go.mod h1:JuG+u74mvjvcm8vj8pI5XiHy1zDeoCS2LB1spIq7Ay0=
 go.opentelemetry.io/otel/exporters/otlp/otlptrace v1.24.0 h1:t6wl9SPayj+c7lEIFgm4ooDBZVb01IhLB4InpomhRw8=
 go.opentelemetry.io/otel/exporters/otlp/otlptrace v1.24.0/go.mod h1:iSDOcsnSA5INXzZtwaBPrKp/lWu/V14Dd+llD0oI2EA=
 go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracehttp v1.24.0 h1:Xw8U6u2f8DK2XAkGRFV7BBLENgnTGX9i4rQRxJf+/vs=
 go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracehttp v1.24.0/go.mod h1:6KW1Fm6R/s6Z3PGXwSJN2K4eT6wQB3vXX6CVnYX9NmM=
-go.opentelemetry.io/otel/metric v1.41.0 h1:rFnDcs4gRzBcsO9tS8LCpgR0dxg4aaxWlJxCno7JlTQ=
-go.opentelemetry.io/otel/metric v1.41.0/go.mod h1:xPvCwd9pU0VN8tPZYzDZV/BMj9CM9vs00GuBjeKhJps=
-go.opentelemetry.io/otel/sdk v1.41.0 h1:YPIEXKmiAwkGl3Gu1huk1aYWwtpRLeskpV+wPisxBp8=
-go.opentelemetry.io/otel/sdk v1.41.0/go.mod h1:ahFdU0G5y8IxglBf0QBJXgSe7agzjE4GiTJ6HT9ud90=
-go.opentelemetry.io/otel/trace v1.41.0 h1:Vbk2co6bhj8L59ZJ6/xFTskY+tGAbOnCtQGVVa9TIN0=
-go.opentelemetry.io/otel/trace v1.41.0/go.mod h1:U1NU4ULCoxeDKc09yCWdWe+3QoyweJcISEVa1RBzOis=
+go.opentelemetry.io/otel/metric v1.43.0 h1:d7638QeInOnuwOONPp4JAOGfbCEpYb+K6DVWvdxGzgM=
+go.opentelemetry.io/otel/metric v1.43.0/go.mod h1:RDnPtIxvqlgO8GRW18W6Z/4P462ldprJtfxHxyKd2PY=
+go.opentelemetry.io/otel/sdk v1.43.0 h1:pi5mE86i5rTeLXqoF/hhiBtUNcrAGHLKQdhg4h4V9Dg=
+go.opentelemetry.io/otel/sdk v1.43.0/go.mod h1:P+IkVU3iWukmiit/Yf9AWvpyRDlUeBaRg6Y+C58QHzg=
+go.opentelemetry.io/otel/sdk/metric v1.43.0 h1:S88dyqXjJkuBNLeMcVPRFXpRw2fuwdvfCGLEo89fDkw=
+go.opentelemetry.io/otel/sdk/metric v1.43.0/go.mod h1:C/RJtwSEJ5hzTiUz5pXF1kILHStzb9zFlIEe85bhj6A=
+go.opentelemetry.io/otel/trace v1.43.0 h1:BkNrHpup+4k4w+ZZ86CZoHHEkohws8AY+WTX09nk+3A=
+go.opentelemetry.io/otel/trace v1.43.0/go.mod h1:/QJhyVBUUswCphDVxq+8mld+AvhXZLhe+8WVFxiFff0=
 go.opentelemetry.io/proto/otlp v1.3.1 h1:TrMUixzpM0yuc/znrFTP9MMRh8trP93mkCiDVeXrui0=
 go.opentelemetry.io/proto/otlp v1.3.1/go.mod h1:0X1WI4de4ZsLrrJNLAQbFeLCm3T7yBkR0XqQ7niQU+8=
 go.uber.org/atomic v1.10.0 h1:9qC72Qh0+3MqyJbAn8YU5xVq1frD8bn3JtD2oXtafVQ=
@@ -608,6 +620,8 @@ golang.org/x/sys v0.0.0-20180830151530-49385e6e1522/go.mod h1:STP8DvDyc/dI5b8T5h
 golang.org/x/sys v0.0.0-20190215142949-d0b11bdaac8a/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
 golang.org/x/sys v0.0.0-20190412213103-97732733099d/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190916202348-b4ddaad3f8a3/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20200116001909-b77594299b42/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20200223170610-d5e6a3e2c0ae/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20200323222414-85ca7c5b95cd/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20200509044756-6aff5f38e54f/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20200930185726-fdedc70b468f/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
@@ -615,6 +629,9 @@ golang.org/x/sys v0.0.0-20201119102817-f84b799fce68/go.mod h1:h1NjWce9XRLGQEsW7w
 golang.org/x/sys v0.0.0-20201204225414-ed752295db88/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20210615035016-665e8c7367d1/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.0.0-20210616094352-59db8d763f22/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/sys v0.0.0-20210630005230-0f9fa26af87c/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/sys v0.0.0-20210927094055-39ccf1dd6fa6/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/sys v0.0.0-20220503163025-988cb79eb6c6/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.0.0-20220520151302-bc2c85ada10a/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.0.0-20220704084225-05e143d24a9e/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.0.0-20220715151400-c0bba94af5f8/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
@@ -678,29 +695,31 @@ golang.org/x/tools v0.47.0/go.mod h1:dFHnyTvFWY212G+h7ZY4Vsp/K3U4/7W9TyVaAul8uCA
 golang.org/x/xerrors v0.0.0-20190717185122-a985d3407aa7/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
 golang.org/x/xerrors v0.0.0-20191011141410-1b5146add898/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
 golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
+gonum.org/v1/gonum v0.17.0 h1:VbpOemQlsSMrYmn7T2OUvQ4dqxQXU+ouZFQsZOx50z4=
+gonum.org/v1/gonum v0.17.0/go.mod h1:El3tOrEuMpv2UdMrbNlKEh9vd86bmQ6vqIcDwxEOc1E=
 google.golang.org/appengine v1.1.0/go.mod h1:EbEs0AVv82hx2wNQdGPgUI5lhzA/G0D9YwlJXL52JkM=
 google.golang.org/appengine v1.4.0/go.mod h1:xpcJRLb0r/rnEns0DIKYYv+WjYCduHsrkT7/EB5XEv4=
 google.golang.org/genproto v0.0.0-20180817151627-c66870c02cf8/go.mod h1:JiN7NxoALGmiZfu7CAH4rXhgtRTLTxftemlI0sWmxmc=
 google.golang.org/genproto v0.0.0-20190819201941-24fa4b261c55/go.mod h1:DMBHOl98Agz4BDEuKkezgsaosCRResVns1a3J2ZsMNc=
 google.golang.org/genproto v0.0.0-20231106174013-bbf56f31fb17 h1:wpZ8pe2x1Q3f2KyT5f8oP/fa9rHAKgFPr/HZdNuS+PQ=
-google.golang.org/genproto/googleapis/api v0.0.0-20250929231259-57b25ae835d4 h1:8XJ4pajGwOlasW+L13MnEGA8W4115jJySQtVfS2/IBU=
-google.golang.org/genproto/googleapis/api v0.0.0-20250929231259-57b25ae835d4/go.mod h1:NnuHhy+bxcg30o7FnVAZbXsPHUDQ9qKWAQKCD7VxFtk=
-google.golang.org/genproto/googleapis/rpc v0.0.0-20250929231259-57b25ae835d4 h1:i8QOKZfYg6AbGVZzUAY3LrNWCKF8O6zFisU9Wl9RER4=
-google.golang.org/genproto/googleapis/rpc v0.0.0-20250929231259-57b25ae835d4/go.mod h1:HSkG/KdJWusxU1F6CNrwNDjBMgisKxGnc5dAZfT0mjQ=
+google.golang.org/genproto/googleapis/api v0.0.0-20260414002931-afd174a4e478 h1:yQugLulqltosq0B/f8l4w9VryjV+N/5gcW0jQ3N8Qec=
+google.golang.org/genproto/googleapis/api v0.0.0-20260414002931-afd174a4e478/go.mod h1:C6ADNqOxbgdUUeRTU+LCHDPB9ttAMCTff6auwCVa4uc=
+google.golang.org/genproto/googleapis/rpc v0.0.0-20260414002931-afd174a4e478 h1:RmoJA1ujG+/lRGNfUnOMfhCy5EipVMyvUE+KNbPbTlw=
+google.golang.org/genproto/googleapis/rpc v0.0.0-20260414002931-afd174a4e478/go.mod h1:4Hqkh8ycfw05ld/3BWL7rJOSfebL2Q+DVDeRgYgxUU8=
 google.golang.org/grpc v1.19.0/go.mod h1:mqu4LbDTu4XGKhr4mRzUsmM4RtVoemTSY81AxZiDr8c=
 google.golang.org/grpc v1.23.0/go.mod h1:Y5yQAOtifL1yxbo5wqy6BxZv8vAUGQwXBOALyacEbxg=
 google.golang.org/grpc v1.25.1/go.mod h1:c3i+UQWmh7LiEpx4sFZnkU36qjEYZ0imhYfXVyQciAY=
 google.golang.org/grpc v1.31.0/go.mod h1:N36X2cJ7JwdamYAgDz+s+rVMFjt3numwzf/HckM8pak=
-google.golang.org/grpc v1.75.1 h1:/ODCNEuf9VghjgO3rqLcfg8fiOP0nSluljWFlDxELLI=
-google.golang.org/grpc v1.75.1/go.mod h1:JtPAzKiq4v1xcAB2hydNlWI2RnF85XXcV0mhKXr2ecQ=
+google.golang.org/grpc v1.82.1 h1:NnAxzGRA0677vCa4BUkOAnO5+FfQqVl9iUXeD0IqcGE=
+google.golang.org/grpc v1.82.1/go.mod h1:yzTZ1TB1Z3SG+LIYaI+WiE8D5+PZ3ArnrSp8zF3+/ZA=
 google.golang.org/protobuf v0.0.0-20200109180630-ec00e32a8dfd/go.mod h1:DFci5gLYBciE7Vtevhsrf46CRTquxDuWsQurQQe4oz8=
 google.golang.org/protobuf v0.0.0-20200221191635-4d8936d0db64/go.mod h1:kwYJMbMJ01Woi6D6+Kah6886xMZcty6N08ah7+eCXa0=
 google.golang.org/protobuf v0.0.0-20200228230310-ab0ca4ff8a60/go.mod h1:cfTl7dwQJ+fmap5saPgwCLgHXTUD7jkjRqWcaiX5VyM=
 google.golang.org/protobuf v1.20.1-0.20200309200217-e05f789c0967/go.mod h1:A+miEFZTKqfCUM6K7xSMQL9OKL/b6hQv+e19PK+JZNE=
 google.golang.org/protobuf v1.21.0/go.mod h1:47Nbq4nVaFHyn7ilMalzfO3qCViNmqZ2kzikPIcrTAo=
 google.golang.org/protobuf v1.23.0/go.mod h1:EGpADcykh3NcUnDUJcl1+ZksZNG86OlYog2l/sGQquU=
-google.golang.org/protobuf v1.36.10 h1:AYd7cD/uASjIL6Q9LiTjz8JLcrh/88q5UObnmY3aOOE=
-google.golang.org/protobuf v1.36.10/go.mod h1:HTf+CrKn2C3g5S8VImy6tdcUvCska2kB7j23XfzDpco=
+google.golang.org/protobuf v1.36.11 h1:fV6ZwhNocDyBLK0dj+fg8ektcVegBBuEolpbTQyBNVE=
+google.golang.org/protobuf v1.36.11/go.mod h1:HTf+CrKn2C3g5S8VImy6tdcUvCska2kB7j23XfzDpco=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
 gopkg.in/check.v1 v1.0.0-20200227125254-8fa46927fb4f/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
 gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c h1:Hei/4ADfdWqJk1ZMxUNpqntNwaWcugrBjAiHlqqRiVk=

backend/internal/config/config.go

@@ -32,7 +32,7 @@ const (
 
 // DefaultCSPPolicy is the default Content-Security-Policy with nonce support
 // __CSP_NONCE__ will be replaced with actual nonce at request time by the SecurityHeaders middleware
-const DefaultCSPPolicy = "default-src 'self'; worker-src 'self' blob:; script-src 'self' __CSP_NONCE__ https://challenges.cloudflare.com https://*.alicdn.com https://static.cloudflareinsights.com https://turing.captcha.qcloud.com https://turing.captcha.gtimg.com https://ca.turing.captcha.qcloud.com https://global.turing.captcha.gtimg.com https://www.tycaptcha.com https://cloudcache.tencentcs.com https://*.stripe.com https://static.airwallex.com https://checkout.airwallex.com https://static-demo.airwallex.com https://checkout-demo.airwallex.com; style-src 'self' 'unsafe-inline' https://*.captcha.gtimg.com https://fonts.googleapis.com https://*.alicdn.com https://static.airwallex.com https://checkout.airwallex.com https://static-demo.airwallex.com https://checkout-demo.airwallex.com; img-src 'self' data: blob: https:; font-src 'self' data: https://fonts.gstatic.com; connect-src 'self' https://turing.captcha.qcloud.com https://www.tycaptcha.com https://rce.tencentrio.com https:; frame-src https://challenges.cloudflare.com https://turing.captcha.qcloud.com https://ca.turing.captcha.qcloud.com https://www.tycaptcha.com https://*.stripe.com https://checkout.airwallex.com https://checkout-demo.airwallex.com; frame-ancestors 'none'; base-uri 'self'; form-action 'self'"
+const DefaultCSPPolicy = "default-src 'self'; worker-src 'self' blob:; script-src 'self' __CSP_NONCE__ https://challenges.cloudflare.com https://*.alicdn.com https://static.cloudflareinsights.com https://turing.captcha.qcloud.com https://turing.captcha.gtimg.com https://ca.turing.captcha.qcloud.com https://global.turing.captcha.gtimg.com https://www.tycaptcha.com https://cloudcache.tencentcs.com https://*.stripe.com https://static.airwallex.com https://checkout.airwallex.com https://static-demo.airwallex.com https://checkout-demo.airwallex.com; style-src 'self' 'unsafe-inline' https://*.captcha.gtimg.com https://fonts.googleapis.com https://*.alicdn.com https://static.airwallex.com https://checkout.airwallex.com https://static-demo.airwallex.com https://checkout-demo.airwallex.com; img-src 'self' data: blob: https:; font-src 'self' data: https://fonts.gstatic.com; connect-src 'self' https://turing.captcha.qcloud.com https://www.tycaptcha.com https://rce.tencentrio.com https:; frame-src 'self' https://challenges.cloudflare.com https://turing.captcha.qcloud.com https://ca.turing.captcha.qcloud.com https://www.tycaptcha.com https://*.stripe.com https://checkout.airwallex.com https://checkout-demo.airwallex.com; frame-ancestors 'none'; base-uri 'self'; form-action 'self'"
 
 // UMQ（用户消息队列）模式常量
 const (
@@ -61,6 +61,10 @@ const (
 // 可通过 gateway.upstream_response_read_max_bytes 配置项覆盖。
 const DefaultUpstreamResponseReadMaxBytes int64 = 128 * 1024 * 1024
 
+// DefaultModelsListReadMaxBytes 上游模型列表响应体的默认读取上限。
+// 可通过 gateway.models_list_read_max_bytes 配置项覆盖。
+const DefaultModelsListReadMaxBytes int64 = 8 * 1024 * 1024
+
 type Config struct {
 	Server                  ServerConfig                  `mapstructure:"server"`
 	Log                     LogConfig                     `mapstructure:"log"`
@@ -99,6 +103,18 @@ type Config struct {
 	Idempotency             IdempotencyConfig             `mapstructure:"idempotency"`
 	BatchImage              BatchImageConfig              `mapstructure:"batch_image"`
 	ImageStorage            ImageStorageConfig            `mapstructure:"image_storage"`
+	Plugins                 PluginConfig                  `mapstructure:"plugins"`
+}
+
+// PluginConfig 控制管理员手动上传的本地进程插件。
+// 默认不包含插件，也不允许安装未签名插件；TrustedPublishers 用于追加第三方发布者。
+type PluginConfig struct {
+	DataDir              string            `mapstructure:"data_dir"`
+	AllowUnsigned        bool              `mapstructure:"allow_unsigned"`
+	TrustedPublishers    map[string]string `mapstructure:"trusted_publishers"`
+	MaxUploadBytes       int64             `mapstructure:"max_upload_bytes"`
+	MaxUncompressedBytes int64             `mapstructure:"max_uncompressed_bytes"`
+	StartTimeoutSeconds  int               `mapstructure:"start_timeout_seconds"`
 }
 
 type LogConfig struct {
@@ -651,6 +667,8 @@ type PricingConfig struct {
 	DataDir string `mapstructure:"data_dir"`
 	// 回退文件路径
 	FallbackFile string `mapstructure:"fallback_file"`
+	// 覆盖补丁文件路径（可选）：条目按字段浅合并覆盖目录/回退数据，优先级最高
+	OverrideFile string `mapstructure:"override_file"`
 	// 更新间隔（小时）
 	UpdateIntervalHours int `mapstructure:"update_interval_hours"`
 	// 哈希校验间隔（分钟）
@@ -830,6 +848,53 @@ type ProxyFallbackConfig struct {
 
 type ProxyProbeConfig struct {
 	InsecureSkipVerify bool `mapstructure:"insecure_skip_verify"` // 已禁用：禁止跳过 TLS 证书验证
+	// URLs 按优先级排列的自定义探测 URL 列表。
+	// 留空时使用内置默认列表（ip-api → ipify）。
+	// 某些 AI API 专用代理只允许访问特定域名，配置多个备选可提高探测成功率。
+	URLs []ProbeURLConfig `mapstructure:"urls"`
+}
+
+// ProbeURLConfig 描述一个探测端点及其响应解析方式。
+type ProbeURLConfig struct {
+	URL    string `mapstructure:"url"`
+	Parser string `mapstructure:"parser"` // "ip-api" / "ipify" / "chatgpt-trace"
+}
+
+func normalizeProxyProbeURLs(targets []ProbeURLConfig) ([]ProbeURLConfig, error) {
+	if len(targets) == 0 {
+		return nil, nil
+	}
+
+	normalized := make([]ProbeURLConfig, 0, len(targets))
+	for i, target := range targets {
+		rawURL := strings.TrimSpace(target.URL)
+		parser := strings.ToLower(strings.TrimSpace(target.Parser))
+		if rawURL == "" {
+			return nil, fmt.Errorf("entry %d: url is required", i)
+		}
+		if parser == "" {
+			return nil, fmt.Errorf("entry %d: parser is required", i)
+		}
+		switch parser {
+		case "ip-api", "ipify", "chatgpt-trace":
+		default:
+			return nil, fmt.Errorf("entry %d: unsupported parser %q", i, target.Parser)
+		}
+
+		parsed, err := url.Parse(rawURL)
+		if err != nil || parsed.Host == "" {
+			return nil, fmt.Errorf("entry %d: invalid url %q", i, target.URL)
+		}
+		if parsed.Scheme != "http" && parsed.Scheme != "https" {
+			return nil, fmt.Errorf("entry %d: url scheme must be http or https", i)
+		}
+
+		normalized = append(normalized, ProbeURLConfig{
+			URL:    rawURL,
+			Parser: parser,
+		})
+	}
+	return normalized, nil
 }
 
 type BillingConfig struct {
@@ -887,6 +952,9 @@ type GatewayConfig struct {
 	// OpenAIResponseHeaderTimeout: OpenAI/Codex 上游等待响应头的超时时间（秒），0表示无超时
 	// OpenAI/Codex 请求可能在上游排队较久；默认不使用通用响应头超时截断。
 	OpenAIResponseHeaderTimeout int `mapstructure:"openai_response_header_timeout"`
+	// GrokResponseHeaderTimeout bounds the pre-first-byte wait for xAI/Grok.
+	// A zero value uses the provider-safe default instead of the generic gateway timeout.
+	GrokResponseHeaderTimeout int `mapstructure:"grok_response_header_timeout"`
 	// OpenAIFirstOutputTimeoutSeconds: native HTTP Responses 首个语义输出超时（秒），0表示禁用。
 	OpenAIFirstOutputTimeoutSeconds int `mapstructure:"openai_first_output_timeout_seconds"`
 	// OpenAIHighEffortFirstOutputTimeoutSeconds: high/xhigh/max 推理的首个语义输出超时（秒）。
@@ -898,6 +966,8 @@ type GatewayConfig struct {
 	TextMaxBodySize int64 `mapstructure:"text_max_body_size"`
 	// 非流式上游响应体读取上限（字节），用于防止无界读取导致内存放大
 	UpstreamResponseReadMaxBytes int64 `mapstructure:"upstream_response_read_max_bytes"`
+	// 上游模型列表响应体读取上限（字节）
+	ModelsListReadMaxBytes int64 `mapstructure:"models_list_read_max_bytes"`
 	// 代理探测响应体读取上限（字节）
 	ProxyProbeResponseReadMaxBytes int64 `mapstructure:"proxy_probe_response_read_max_bytes"`
 	// Gemini 上游响应头调试日志开关（默认关闭，避免高频日志开销）
@@ -2211,14 +2281,23 @@ func setDefaults() {
 	viper.SetDefault("rate_limit.overload_cooldown_minutes", 10)
 	viper.SetDefault("rate_limit.oauth_401_cooldown_minutes", 10)
 
-	// Pricing - 从 model-price-repo 同步模型定价和上下文窗口数据（固定到 commit，避免分支漂移）
+	// Pricing - 从 model-price-repo main 分支同步模型定价和上下文窗口数据
 	viper.SetDefault("pricing.remote_url", "https://raw.githubusercontent.com/Wei-Shaw/model-price-repo/main/model_prices_and_context_window.json")
 	viper.SetDefault("pricing.hash_url", "https://raw.githubusercontent.com/Wei-Shaw/model-price-repo/main/model_prices_and_context_window.sha256")
 	viper.SetDefault("pricing.data_dir", "./data")
 	viper.SetDefault("pricing.fallback_file", "./resources/model-pricing/model_prices_and_context_window.json")
+	viper.SetDefault("pricing.override_file", "")
 	viper.SetDefault("pricing.update_interval_hours", 24)
 	viper.SetDefault("pricing.hash_check_interval_minutes", 10)
 
+	// 本地进程插件。插件必须由管理员手动上传，项目默认不携带任何插件能力。
+	viper.SetDefault("plugins.data_dir", "")
+	viper.SetDefault("plugins.allow_unsigned", false)
+	viper.SetDefault("plugins.trusted_publishers", map[string]string{})
+	viper.SetDefault("plugins.max_upload_bytes", int64(128*1024*1024))
+	viper.SetDefault("plugins.max_uncompressed_bytes", int64(256*1024*1024))
+	viper.SetDefault("plugins.start_timeout_seconds", 15)
+
 	// Timezone (default to Asia/Shanghai for Chinese users)
 	viper.SetDefault("timezone", "Asia/Shanghai")
 
@@ -2280,6 +2359,7 @@ func setDefaults() {
 	// Gateway
 	viper.SetDefault("gateway.response_header_timeout", 600) // 600秒(10分钟)等待上游响应头，LLM高负载时可能排队较久
 	viper.SetDefault("gateway.openai_response_header_timeout", 0)
+	viper.SetDefault("gateway.grok_response_header_timeout", 120)
 	viper.SetDefault("gateway.openai_first_output_timeout_seconds", 0)
 	viper.SetDefault("gateway.openai_high_effort_first_output_timeout_seconds", 0)
 	viper.SetDefault("gateway.log_upstream_error_body", true)
@@ -2385,6 +2465,7 @@ func setDefaults() {
 	viper.SetDefault("gateway.max_body_size", int64(256*1024*1024))
 	viper.SetDefault("gateway.text_max_body_size", int64(32*1024*1024))
 	viper.SetDefault("gateway.upstream_response_read_max_bytes", DefaultUpstreamResponseReadMaxBytes)
+	viper.SetDefault("gateway.models_list_read_max_bytes", DefaultModelsListReadMaxBytes)
 	viper.SetDefault("gateway.proxy_probe_response_read_max_bytes", int64(1024*1024))
 	viper.SetDefault("gateway.gemini_debug_response_headers", false)
 	viper.SetDefault("gateway.connection_pool_isolation", ConnectionPoolIsolationAccountProxy)
@@ -2569,6 +2650,20 @@ func (c *Config) Validate() error {
 	}
 	c.Security.ForwardedClientIPHeaders = forwardedClientIPHeaders
 	c.SetForwardedClientIPSettings(c.Security.TrustForwardedIPForAPIKeyACL, forwardedClientIPHeaders)
+	proxyProbeURLs, err := normalizeProxyProbeURLs(c.Security.ProxyProbe.URLs)
+	if err != nil {
+		return fmt.Errorf("security.proxy_probe.urls: %w", err)
+	}
+	c.Security.ProxyProbe.URLs = proxyProbeURLs
+	if c.Plugins.MaxUploadBytes <= 0 || c.Plugins.MaxUploadBytes > 1024*1024*1024 {
+		return fmt.Errorf("plugins.max_upload_bytes must be between 1 and 1073741824")
+	}
+	if c.Plugins.MaxUncompressedBytes < c.Plugins.MaxUploadBytes || c.Plugins.MaxUncompressedBytes > 2*1024*1024*1024 {
+		return fmt.Errorf("plugins.max_uncompressed_bytes must be between max_upload_bytes and 2147483648")
+	}
+	if c.Plugins.StartTimeoutSeconds < 1 || c.Plugins.StartTimeoutSeconds > 120 {
+		return fmt.Errorf("plugins.start_timeout_seconds must be between 1 and 120")
+	}
 	if c.Server.ReadHeaderTimeout < 1 || c.Server.ReadHeaderTimeout > 60 {
 		return fmt.Errorf("server.read_header_timeout must be between 1 and 60 seconds")
 	}
@@ -3180,6 +3275,9 @@ func (c *Config) Validate() error {
 	if c.Gateway.UpstreamResponseReadMaxBytes <= 0 {
 		return fmt.Errorf("gateway.upstream_response_read_max_bytes must be positive")
 	}
+	if c.Gateway.ModelsListReadMaxBytes <= 0 {
+		return fmt.Errorf("gateway.models_list_read_max_bytes must be positive")
+	}
 	if c.Gateway.ProxyProbeResponseReadMaxBytes <= 0 {
 		return fmt.Errorf("gateway.proxy_probe_response_read_max_bytes must be positive")
 	}
@@ -3189,6 +3287,9 @@ func (c *Config) Validate() error {
 	if c.Gateway.OpenAIResponseHeaderTimeout < 0 {
 		return fmt.Errorf("gateway.openai_response_header_timeout must be non-negative")
 	}
+	if c.Gateway.GrokResponseHeaderTimeout < 0 || c.Gateway.GrokResponseHeaderTimeout > 1800 {
+		return fmt.Errorf("gateway.grok_response_header_timeout must be between 0-1800 seconds")
+	}
 	if c.Gateway.OpenAIFirstOutputTimeoutSeconds < 0 || c.Gateway.OpenAIFirstOutputTimeoutSeconds > 600 ||
 		(c.Gateway.OpenAIFirstOutputTimeoutSeconds > 0 && c.Gateway.OpenAIFirstOutputTimeoutSeconds < 30) {
 		return fmt.Errorf("gateway.openai_first_output_timeout_seconds must be 0 or between 30-600 seconds")

backend/internal/config/config_test.go

@@ -23,6 +23,13 @@ func resetViperWithJWTSecret(t *testing.T) {
 	t.Setenv("JWT_SECRET", strings.Repeat("x", 32))
 }
 
+func TestLoadDefaultModelsListReadMaxBytes(t *testing.T) {
+	resetViperWithJWTSecret(t)
+	cfg, err := Load()
+	require.NoError(t, err)
+	require.Equal(t, DefaultModelsListReadMaxBytes, cfg.Gateway.ModelsListReadMaxBytes)
+}
+
 func TestLoadTimezonePrecedence(t *testing.T) {
 	tests := []struct {
 		name         string
@@ -553,6 +560,15 @@ func TestLoadOpenAIWSClientFirstMessageTimeoutFromEnv(t *testing.T) {
 	require.Equal(t, 120, cfg.Gateway.OpenAIWS.ClientFirstMessageTimeoutSeconds)
 }
 
+func TestLoadOpenAIWSForceHTTPFromEnv(t *testing.T) {
+	resetViperWithJWTSecret(t)
+	t.Setenv("GATEWAY_OPENAI_WS_FORCE_HTTP", "true")
+
+	cfg, err := Load()
+	require.NoError(t, err)
+	require.True(t, cfg.Gateway.OpenAIWS.ForceHTTP)
+}
+
 func TestLoadDefaultOpenAICompactModel(t *testing.T) {
 	resetViperWithJWTSecret(t)
 
@@ -1794,6 +1810,11 @@ func TestValidateConfigErrors(t *testing.T) {
 			mutate:  func(c *Config) { c.Gateway.TextMaxBodySize = c.Gateway.MaxBodySize + 1 },
 			wantErr: "gateway.text_max_body_size",
 		},
+		{
+			name:    "gateway models list read limit",
+			mutate:  func(c *Config) { c.Gateway.ModelsListReadMaxBytes = 0 },
+			wantErr: "gateway.models_list_read_max_bytes",
+		},
 		{
 			name:    "gateway response header timeout",
 			mutate:  func(c *Config) { c.Gateway.ResponseHeaderTimeout = -1 },

backend/internal/config/env_reachability_test.go

@@ -45,6 +45,20 @@ func collectMapstructureKeys(t reflect.Type, prefix string, out map[string]strin
 			// is out of scope here — such settings need a config file either way.
 			continue
 		}
+		if ft.Kind() == reflect.Slice {
+			elem := ft.Elem()
+			for elem.Kind() == reflect.Ptr {
+				elem = elem.Elem()
+			}
+			if elem.Kind() == reflect.Struct {
+				// AutomaticEnv exposes one string value. Viper's string-to-slice
+				// hook can populate scalar slices, but it cannot decode a string
+				// into []struct. Registering a default would turn silent ignore
+				// into a startup unmarshal error, so structured slices remain
+				// config-file-only just like maps.
+				continue
+			}
+		}
 		out[strings.ToLower(key)] = ft.String()
 	}
 }
@@ -62,7 +76,7 @@ func collectMapstructureKeys(t reflect.Type, prefix string, out map[string]strin
 // were lost, silently disabling async image tasks for env-driven deployments.
 //
 // When this fails, register a zero-valued default in setEnvReachableDefaults
-// for each reported key.
+// for each reported scalar key. Maps and slices of structs are config-file-only.
 func TestConfigKeysAreEnvReachable(t *testing.T) {
 	bound := map[string]string{}
 	collectMapstructureKeys(reflect.TypeOf(Config{}), "", bound)

backend/internal/config/proxy_probe_config_test.go

@@ -0,0 +1,47 @@
+//go:build unit
+
+package config
+
+import (
+	"testing"
+
+	"github.com/stretchr/testify/require"
+)
+
+func TestNormalizeProxyProbeURLs(t *testing.T) {
+	t.Parallel()
+
+	got, err := normalizeProxyProbeURLs([]ProbeURLConfig{
+		{URL: " https://chatgpt.com/cdn-cgi/trace ", Parser: " CHATGPT-TRACE "},
+		{URL: "https://api64.ipify.org?format=json", Parser: "ipify"},
+	})
+	require.NoError(t, err)
+	require.Equal(t, []ProbeURLConfig{
+		{URL: "https://chatgpt.com/cdn-cgi/trace", Parser: "chatgpt-trace"},
+		{URL: "https://api64.ipify.org?format=json", Parser: "ipify"},
+	}, got)
+}
+
+func TestNormalizeProxyProbeURLsRejectsInvalidEntries(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		name    string
+		target  ProbeURLConfig
+		wantErr string
+	}{
+		{name: "missing URL", target: ProbeURLConfig{Parser: "ipify"}, wantErr: "url is required"},
+		{name: "missing parser", target: ProbeURLConfig{URL: "https://example.com"}, wantErr: "parser is required"},
+		{name: "unknown parser", target: ProbeURLConfig{URL: "https://example.com", Parser: "ip_api"}, wantErr: "unsupported parser"},
+		{name: "relative URL", target: ProbeURLConfig{URL: "/cdn-cgi/trace", Parser: "chatgpt-trace"}, wantErr: "invalid url"},
+		{name: "unsupported scheme", target: ProbeURLConfig{URL: "ftp://example.com/file", Parser: "ipify"}, wantErr: "scheme must be http or https"},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+			_, err := normalizeProxyProbeURLs([]ProbeURLConfig{tt.target})
+			require.ErrorContains(t, err, tt.wantErr)
+		})
+	}
+}

backend/internal/domain/channel_monitor_quota.go

@@ -50,6 +50,10 @@ type MonitorQuotaSnapshot struct {
 	Balances  []MonitorBalance   `json:"balances,omitempty"`   // 多币种余额（如 DeepSeek CNY+USD）
 	Currency  string             `json:"currency,omitempty"`   // 主余额币种
 	PlanLevel string             `json:"plan_level,omitempty"` // 套餐等级（如智谱 level）
+	// BalanceLow 余额低于阈值或账号被上游标记不可用（仅 cn_balance 来源）。
+	// 抓取器按 Gateway.CNProviders.BalanceThreshold 判定，口径与账号停调
+	// （CNProviderBalanceCheckService.checkOne）一致：任一币种达标即健康。
+	BalanceLow bool `json:"balance_low,omitempty"`
 	// CredentialInvalid 上游 401/403 鉴权失败（区别于网络/解析错误），
 	// 检测状态据此推导 failed 而非 error。
 	CredentialInvalid bool      `json:"credential_invalid,omitempty"`

backend/internal/domain/constants.go

@@ -44,6 +44,7 @@ const (
 	APIProtocolChatCompletions = "chat_completions" // OpenAI Chat Completions（默认）
 	APIProtocolAnthropic       = "anthropic"        // 原生 Anthropic /v1/messages（适配 Claude Code）
 	APIProtocolResponses       = "responses"        // OpenAI Responses（仅 deepseek，适配 Codex）
+	APIProtocolAdaptive        = "adaptive"         // 按入站协议优先选择供应商原生端点
 )
 
 // Account type constants
@@ -104,11 +105,11 @@ var DefaultAntigravityModelMapping = map[string]string{
 	"claude-opus-4-6":            "claude-opus-4-6-thinking", // 简称映射
 	"claude-opus-4-5-thinking":   "claude-opus-4-6-thinking", // 迁移旧模型
 	"claude-sonnet-4-6":          "claude-sonnet-4-6",
-	"claude-sonnet-4-5":          "claude-sonnet-4-5",
-	"claude-sonnet-4-5-thinking": "claude-sonnet-4-5-thinking",
+	"claude-sonnet-4-5":          "claude-sonnet-4-5", // 显式 canonical 选择透传
+	"claude-sonnet-4-5-thinking": "claude-sonnet-4-6", // 迁移旧兼容别名
 	// Claude 详细版本 ID 映射
 	"claude-opus-4-5-20251101":   "claude-opus-4-6-thinking", // 迁移旧模型
-	"claude-sonnet-4-5-20250929": "claude-sonnet-4-5",
+	"claude-sonnet-4-5-20250929": "claude-sonnet-4-6",        // 迁移旧模型
 	// Claude Haiku → Sonnet（无 Haiku 支持）
 	"claude-haiku-4-5":          "claude-sonnet-4-6",
 	"claude-haiku-4-5-20251001": "claude-sonnet-4-6",

backend/internal/domain/constants_test.go

@@ -43,6 +43,21 @@ func TestDefaultAntigravityModelMapping_ContainsNewClaudeModels(t *testing.T) {
 	}
 }
 
+func TestDefaultAntigravityModelMapping_PreservesExplicitSonnet45AndMigratesLegacyAliases(t *testing.T) {
+	t.Parallel()
+
+	cases := map[string]string{
+		"claude-sonnet-4-5":          "claude-sonnet-4-5",
+		"claude-sonnet-4-5-thinking": "claude-sonnet-4-6",
+		"claude-sonnet-4-5-20250929": "claude-sonnet-4-6",
+	}
+	for model, want := range cases {
+		if got := DefaultAntigravityModelMapping[model]; got != want {
+			t.Fatalf("expected model %q to map to %q, got %q", model, want, got)
+		}
+	}
+}
+
 func TestDefaultAntigravityModelMapping_Gemini31ProAliases(t *testing.T) {
 	t.Parallel()
 

backend/internal/handler/admin/account_handler.go

@@ -2614,8 +2614,14 @@ func (h *AccountHandler) GetAvailableModels(c *gin.Context) {
 
 	// Handle Gemini accounts
 	if account.IsGemini() {
-		// For OAuth accounts: return default Gemini models
+		// Consumer Google One OAuth still uses the legacy Gemini CLI / Code
+		// Assist channel. Do not advertise newer 3.x or image models that the
+		// channel cannot serve.
 		if account.IsOAuth() {
+			if account.IsGeminiGoogleOne() {
+				response.Success(c, geminicli.GoogleOneModels)
+				return
+			}
 			response.Success(c, geminicli.DefaultModels)
 			return
 		}
@@ -2768,13 +2774,15 @@ func (h *AccountHandler) SyncUpstreamModels(c *gin.Context) {
 		return
 	}
 
-	models, err := h.accountTestService.FetchUpstreamSupportedModels(c.Request.Context(), account)
+	catalog, err := h.accountTestService.SyncUpstreamModelCatalog(c.Request.Context(), account)
 	if err != nil {
 		var syncErr *service.UpstreamModelSyncError
 		if errors.As(err, &syncErr) {
 			switch syncErr.Kind {
 			case service.UpstreamModelSyncErrorConfiguration, service.UpstreamModelSyncErrorUnsupported:
 				response.BadRequest(c, syncErr.SafeMessage())
+			case service.UpstreamModelSyncErrorInternal:
+				response.InternalError(c, syncErr.SafeMessage())
 			default:
 				slog.Warn("sync_upstream_models_failed", "account_id", accountID, "kind", syncErr.Kind)
 				response.Error(c, http.StatusBadGateway, syncErr.SafeMessage())
@@ -2787,29 +2795,35 @@ func (h *AccountHandler) SyncUpstreamModels(c *gin.Context) {
 		return
 	}
 
-	response.Success(c, gin.H{"models": models})
+	response.Success(c, catalog)
 }
 
 // SyncUpstreamModelsPreview handles syncing live supported models using provided credentials (no account ID needed).
 // POST /api/v1/admin/accounts/models/sync-upstream-preview
 func (h *AccountHandler) SyncUpstreamModelsPreview(c *gin.Context) {
 	var req struct {
-		Platform string `json:"platform" binding:"required"`
-		Type     string `json:"type" binding:"required"`
-		BaseURL  string `json:"base_url"`
-		APIKey   string `json:"api_key" binding:"required"`
+		Platform     string            `json:"platform" binding:"required"`
+		Type         string            `json:"type" binding:"required"`
+		BaseURL      string            `json:"base_url"`
+		APIKey       string            `json:"api_key" binding:"required"`
+		ModelMapping map[string]string `json:"model_mapping"`
 	}
 	if err := c.ShouldBindJSON(&req); err != nil {
 		response.BadRequest(c, "Invalid request: "+err.Error())
 		return
 	}
+	modelMapping := make(map[string]any, len(req.ModelMapping))
+	for sourceModel, upstreamModel := range req.ModelMapping {
+		modelMapping[sourceModel] = upstreamModel
+	}
 
 	tempAccount := &service.Account{
 		Platform: req.Platform,
 		Type:     req.Type,
 		Credentials: map[string]any{
-			"api_key":  req.APIKey,
-			"base_url": req.BaseURL,
+			"api_key":       req.APIKey,
+			"base_url":      req.BaseURL,
+			"model_mapping": modelMapping,
 		},
 	}
 
@@ -2818,13 +2832,15 @@ func (h *AccountHandler) SyncUpstreamModelsPreview(c *gin.Context) {
 		return
 	}
 
-	models, err := h.accountTestService.FetchUpstreamSupportedModels(c.Request.Context(), tempAccount)
+	catalog, err := h.accountTestService.SyncUpstreamModelCatalog(c.Request.Context(), tempAccount)
 	if err != nil {
 		var syncErr *service.UpstreamModelSyncError
 		if errors.As(err, &syncErr) {
 			switch syncErr.Kind {
 			case service.UpstreamModelSyncErrorConfiguration, service.UpstreamModelSyncErrorUnsupported:
 				response.BadRequest(c, syncErr.SafeMessage())
+			case service.UpstreamModelSyncErrorInternal:
+				response.InternalError(c, syncErr.SafeMessage())
 			default:
 				slog.Warn("sync_upstream_models_preview_failed", "platform", req.Platform, "kind", syncErr.Kind)
 				response.Error(c, http.StatusBadGateway, syncErr.SafeMessage())
@@ -2837,7 +2853,7 @@ func (h *AccountHandler) SyncUpstreamModelsPreview(c *gin.Context) {
 		return
 	}
 
-	response.Success(c, gin.H{"models": models})
+	response.Success(c, catalog)
 }
 
 // SetPrivacy handles setting privacy for a single OpenAI/Antigravity OAuth account

backend/internal/handler/admin/account_handler_available_models_test.go

@@ -38,14 +38,20 @@ func setupAvailableModelsRouter(adminSvc service.AdminService) *gin.Engine {
 }
 
 type syncUpstreamHTTPUpstream struct {
-	resp *http.Response
-	err  error
+	resp      *http.Response
+	responses []*http.Response
+	err       error
 }
 
 func (u *syncUpstreamHTTPUpstream) Do(req *http.Request, proxyURL string, accountID int64, accountConcurrency int) (*http.Response, error) {
 	if u.err != nil {
 		return nil, u.err
 	}
+	if len(u.responses) > 0 {
+		resp := u.responses[0]
+		u.responses = u.responses[1:]
+		return resp, nil
+	}
 	return u.resp, nil
 }
 
@@ -68,6 +74,7 @@ func setupSyncUpstreamModelsRouter(adminSvc service.AdminService, upstream servi
 	)
 	handler := NewAccountHandler(adminSvc, nil, nil, nil, nil, nil, nil, nil, accountTestSvc, nil, nil, nil, nil, nil)
 	router.POST("/api/v1/admin/accounts/:id/models/sync-upstream", handler.SyncUpstreamModels)
+	router.POST("/api/v1/admin/accounts/models/sync-upstream-preview", handler.SyncUpstreamModelsPreview)
 	return router
 }
 
@@ -287,6 +294,42 @@ func TestAccountHandlerGetAvailableModels_OpenAISparkShadowReturnsMappingModels(
 	}, ids, "影子可用模型由 model_mapping 派生（非写死）")
 }
 
+func TestAccountHandlerGetAvailableModels_GeminiGoogleOneUsesConservativeCatalog(t *testing.T) {
+	svc := &availableModelsAdminService{
+		stubAdminService: newStubAdminService(),
+		account: service.Account{
+			ID:       45,
+			Name:     "google-one",
+			Platform: service.PlatformGemini,
+			Type:     service.AccountTypeOAuth,
+			Status:   service.StatusActive,
+			Credentials: map[string]any{
+				"oauth_type": "google_one",
+			},
+		},
+	}
+	router := setupAvailableModelsRouter(svc)
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodGet, "/api/v1/admin/accounts/45/models", nil)
+	router.ServeHTTP(rec, req)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	var resp struct {
+		Data []struct {
+			ID string `json:"id"`
+		} `json:"data"`
+	}
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &resp))
+	ids := make([]string, 0, len(resp.Data))
+	for _, model := range resp.Data {
+		ids = append(ids, model.ID)
+	}
+	require.ElementsMatch(t, []string{"gemini-2.0-flash", "gemini-2.5-flash", "gemini-2.5-pro"}, ids)
+	require.NotContains(t, ids, "gemini-3.5-flash")
+	require.NotContains(t, ids, "gemini-2.5-flash-image")
+}
+
 func TestAccountHandlerSyncUpstreamModels_ConfigErrorReturnsBadRequest(t *testing.T) {
 	svc := &availableModelsAdminService{
 		stubAdminService: newStubAdminService(),
@@ -311,6 +354,99 @@ func TestAccountHandlerSyncUpstreamModels_ConfigErrorReturnsBadRequest(t *testin
 	require.Contains(t, rec.Body.String(), "No OpenAI API key is available")
 }
 
+func TestAccountHandlerSyncUpstreamModelsReturnsCapabilityMetadata(t *testing.T) {
+	svc := &availableModelsAdminService{
+		stubAdminService: newStubAdminService(),
+		account: service.Account{
+			ID: 48, Name: "custom-openai", Platform: service.PlatformOpenAI,
+			Type: service.AccountTypeAPIKey, Status: service.StatusActive,
+			Credentials: map[string]any{"api_key": "key", "base_url": "https://provider.example/v1"},
+		},
+	}
+	upstream := &syncUpstreamHTTPUpstream{resp: &http.Response{
+		StatusCode: http.StatusOK,
+		Header:     http.Header{"Content-Type": []string{"application/json"}},
+		Body: io.NopCloser(strings.NewReader(`{"models":[{
+			"id":"custom-thinking-model",
+			"reasoning":true,
+			"default_reasoning_level":"high",
+			"supported_reasoning_levels":["low","high"],
+			"input_modalities":["text","image"],
+			"context_window":256000
+		}]}`)),
+	}}
+	router := setupSyncUpstreamModelsRouter(svc, upstream)
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/api/v1/admin/accounts/48/models/sync-upstream", nil)
+	router.ServeHTTP(rec, req)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	var resp struct {
+		Data service.UpstreamModelCatalog `json:"data"`
+	}
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &resp))
+	require.Equal(t, []string{"custom-thinking-model"}, resp.Data.Models)
+	metadata := resp.Data.Metadata["custom-thinking-model"]
+	require.NotNil(t, metadata.Reasoning)
+	require.True(t, *metadata.Reasoning)
+	require.Equal(t, []string{"low", "high"}, metadata.SupportedReasoningLevels)
+	require.Equal(t, []string{"text", "image"}, metadata.InputModalities)
+}
+
+// Scenario: 创建账号 preview 将具体 mapping 传给 404/405 配置回退。
+func TestAccountHandlerSyncUpstreamModelsPreviewUsesProvidedModelMapping(t *testing.T) {
+	upstream := &syncUpstreamHTTPUpstream{responses: []*http.Response{
+		{
+			StatusCode: http.StatusNotFound,
+			Header:     http.Header{"Content-Type": []string{"application/json"}},
+			Body:       io.NopCloser(strings.NewReader(`{"error":"not found"}`)),
+		},
+		{
+			StatusCode: http.StatusOK,
+			Header:     http.Header{"Content-Type": []string{"application/json"}},
+			Body: io.NopCloser(strings.NewReader(`{
+				"configured-provider": {
+					"api": "https://provider.example/v1",
+					"models": {
+						"glm-5.3": {
+							"id": "glm-5.3",
+							"reasoning": true,
+							"reasoning_options": [{"type":"effort","values":["low","high"]}],
+							"modalities": {"input":["text"],"output":["text"]},
+							"limit": {"context":1000000,"output":131072}
+						}
+					}
+				}
+			}`)),
+		},
+	}}
+	router := setupSyncUpstreamModelsRouter(newStubAdminService(), upstream)
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequest(
+		http.MethodPost,
+		"/api/v1/admin/accounts/models/sync-upstream-preview",
+		strings.NewReader(`{
+			"platform":"openai",
+			"type":"apikey",
+			"base_url":"https://provider.example/v1",
+			"api_key":"key",
+			"model_mapping":{"public-glm":"glm-5.3"}
+		}`),
+	)
+	req.Header.Set("Content-Type", "application/json")
+	router.ServeHTTP(rec, req)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	var resp struct {
+		Data service.UpstreamModelCatalog `json:"data"`
+	}
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &resp))
+	require.Equal(t, []string{"glm-5.3"}, resp.Data.Models)
+	require.Equal(t, []string{"low", "high"}, resp.Data.Metadata["glm-5.3"].SupportedReasoningLevels)
+}
+
 func TestAccountHandlerSyncUpstreamModels_UpstreamErrorDoesNotExposeBody(t *testing.T) {
 	svc := &availableModelsAdminService{
 		stubAdminService: newStubAdminService(),
@@ -341,3 +477,52 @@ func TestAccountHandlerSyncUpstreamModels_UpstreamErrorDoesNotExposeBody(t *test
 	require.Contains(t, rec.Body.String(), "Upstream model list request failed with HTTP 502")
 	require.NotContains(t, rec.Body.String(), "SECRET_TOKEN")
 }
+
+// Scenario: 能力补全失败显示部分成功。
+func TestAccountHandlerSyncUpstreamModels_MetadataEnrichmentFailureReturnsWarning(t *testing.T) {
+	svc := &availableModelsAdminService{
+		stubAdminService: newStubAdminService(),
+		account: service.Account{
+			ID:       46,
+			Name:     "opencode-id-only-model-list",
+			Platform: service.PlatformOpenAI,
+			Type:     service.AccountTypeAPIKey,
+			Status:   service.StatusActive,
+			Credentials: map[string]any{
+				"api_key":  "opencode-key",
+				"base_url": "https://opencode.ai/zen/v1",
+			},
+		},
+	}
+	upstream := &syncUpstreamHTTPUpstream{responses: []*http.Response{
+		{
+			StatusCode: http.StatusOK,
+			Header:     http.Header{"Content-Type": []string{"application/json"}},
+			Body:       io.NopCloser(strings.NewReader(`{"data":[{"id":"x-preview-f-free"}]}`)),
+		},
+		{
+			StatusCode: http.StatusBadGateway,
+			Header:     http.Header{"Content-Type": []string{"application/json"}},
+			Body:       io.NopCloser(strings.NewReader(`{"error":"registry unavailable"}`)),
+		},
+	}}
+	router := setupSyncUpstreamModelsRouter(svc, upstream)
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/api/v1/admin/accounts/46/models/sync-upstream", nil)
+	router.ServeHTTP(rec, req)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	var resp struct {
+		Data struct {
+			Models   []string `json:"models"`
+			Warnings []struct {
+				Code string `json:"code"`
+			} `json:"warnings"`
+		} `json:"data"`
+	}
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &resp))
+	require.Equal(t, []string{"x-preview-f-free"}, resp.Data.Models)
+	require.Len(t, resp.Data.Warnings, 1)
+	require.Equal(t, "upstream_model_metadata_incomplete", resp.Data.Warnings[0].Code)
+}

backend/internal/handler/admin/account_upstream_billing_rates.go

@@ -0,0 +1,99 @@
+package admin
+
+import (
+	"crypto/sha256"
+	"encoding/hex"
+	"encoding/json"
+	"net/http"
+	"strconv"
+	"strings"
+
+	"github.com/Wei-Shaw/sub2api/internal/pkg/response"
+	"github.com/Wei-Shaw/sub2api/internal/service"
+	"github.com/gin-gonic/gin"
+)
+
+// upstreamBillingRatesResponse is deliberately smaller than the regular
+// account-list response. The frontend uses the returned ID order to detect a
+// page boundary/order change and only applies snapshots when the page is
+// otherwise stable.
+type upstreamBillingRatesResponse struct {
+	Items    []service.UpstreamBillingRateSnapshotItem `json:"items"`
+	Total    int64                                     `json:"total"`
+	Page     int                                       `json:"page"`
+	PageSize int                                       `json:"page_size"`
+}
+
+// GetUpstreamBillingRates returns persisted upstream billing snapshots for the
+// current account-list page. It never probes an upstream and never loads
+// today-stats, usage, concurrency, or credentials for the response.
+//
+// GET /api/v1/admin/accounts/upstream-billing-rates
+func (h *AccountHandler) GetUpstreamBillingRates(c *gin.Context) {
+	if h.adminService == nil {
+		response.Error(c, http.StatusServiceUnavailable, "account service unavailable")
+		return
+	}
+
+	page, pageSize := response.ParsePagination(c)
+	platform := c.Query("platform")
+	accountType := c.Query("type")
+	status := c.Query("status")
+	search := strings.TrimSpace(c.Query("search"))
+	if len(search) > 100 {
+		search = search[:100]
+	}
+	privacyMode := strings.TrimSpace(c.Query("privacy_mode"))
+	sortBy := c.DefaultQuery("sort_by", "name")
+	sortOrder := c.DefaultQuery("sort_order", "asc")
+
+	var groupID int64
+	if groupQuery := c.Query("group"); groupQuery != "" {
+		if groupQuery == accountListGroupUngroupedQueryValue {
+			groupID = service.AccountListGroupUngrouped
+		} else {
+			parsed, err := strconv.ParseInt(groupQuery, 10, 64)
+			if err != nil || parsed < 0 {
+				response.BadRequest(c, "invalid group filter")
+				return
+			}
+			groupID = parsed
+		}
+	}
+
+	accounts, total, err := h.adminService.ListAccounts(
+		c.Request.Context(), page, pageSize, platform, accountType, status,
+		search, groupID, privacyMode, sortBy, sortOrder,
+	)
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+
+	payload := upstreamBillingRatesResponse{
+		Items:    service.BuildUpstreamBillingRateSnapshotItems(accounts),
+		Total:    total,
+		Page:     page,
+		PageSize: pageSize,
+	}
+	c.Header("Cache-Control", "private, no-cache")
+	etag := buildUpstreamBillingRatesETag(payload)
+	if etag != "" {
+		c.Header("ETag", etag)
+		if ifNoneMatchMatched(c.GetHeader("If-None-Match"), etag) {
+			c.Status(http.StatusNotModified)
+			return
+		}
+	}
+
+	response.Success(c, payload)
+}
+
+func buildUpstreamBillingRatesETag(payload upstreamBillingRatesResponse) string {
+	raw, err := json.Marshal(payload)
+	if err != nil {
+		return ""
+	}
+	sum := sha256.Sum256(raw)
+	return "\"" + hex.EncodeToString(sum[:]) + "\""
+}

backend/internal/handler/admin/channel_handler.go

@@ -64,6 +64,8 @@ type channelModelPricingRequest struct {
 	OutputPrice      *float64                   `json:"output_price" binding:"omitempty,min=0"`
 	CacheWritePrice  *float64                   `json:"cache_write_price" binding:"omitempty,min=0"`
 	CacheReadPrice   *float64                   `json:"cache_read_price" binding:"omitempty,min=0"`
+	FastMultiplier   *float64                   `json:"fast_multiplier" binding:"omitempty,gt=0"`
+	FlexMultiplier   *float64                   `json:"flex_multiplier" binding:"omitempty,gt=0"`
 	ImageInputPrice  *float64                   `json:"image_input_price" binding:"omitempty,min=0"`
 	ImageOutputPrice *float64                   `json:"image_output_price" binding:"omitempty,min=0"`
 	PerRequestPrice  *float64                   `json:"per_request_price" binding:"omitempty,min=0"`
@@ -72,8 +74,9 @@ type channelModelPricingRequest struct {
 }
 
 type channelTimePricingRequest struct {
-	Timezone string                            `json:"timezone"`
-	Periods  []channelTimePricingPeriodRequest `json:"periods"`
+	Timezone     string                            `json:"timezone"`
+	WeekdaysOnly bool                              `json:"weekdays_only"`
+	Periods      []channelTimePricingPeriodRequest `json:"periods"`
 }
 
 type channelTimePricingPeriodRequest struct {
@@ -83,15 +86,19 @@ type channelTimePricingPeriodRequest struct {
 }
 
 type pricingIntervalRequest struct {
-	MinTokens       int      `json:"min_tokens"`
-	MaxTokens       *int     `json:"max_tokens"`
-	TierLabel       string   `json:"tier_label"`
-	InputPrice      *float64 `json:"input_price"`
-	OutputPrice     *float64 `json:"output_price"`
-	CacheWritePrice *float64 `json:"cache_write_price"`
-	CacheReadPrice  *float64 `json:"cache_read_price"`
-	PerRequestPrice *float64 `json:"per_request_price"`
-	SortOrder       int      `json:"sort_order"`
+	MinTokens            int      `json:"min_tokens"`
+	MaxTokens            *int     `json:"max_tokens"`
+	TierLabel            string   `json:"tier_label"`
+	InputPrice           *float64 `json:"input_price"`
+	OutputPrice          *float64 `json:"output_price"`
+	CacheWritePrice      *float64 `json:"cache_write_price"`
+	CacheReadPrice       *float64 `json:"cache_read_price"`
+	InputMultiplier      *float64 `json:"input_multiplier" binding:"omitempty,gt=0"`
+	OutputMultiplier     *float64 `json:"output_multiplier" binding:"omitempty,gt=0"`
+	CacheWriteMultiplier *float64 `json:"cache_write_multiplier" binding:"omitempty,gt=0"`
+	CacheReadMultiplier  *float64 `json:"cache_read_multiplier" binding:"omitempty,gt=0"`
+	PerRequestPrice      *float64 `json:"per_request_price"`
+	SortOrder            int      `json:"sort_order"`
 }
 
 type accountStatsPricingRuleRequest struct {
@@ -128,6 +135,8 @@ type channelModelPricingResponse struct {
 	OutputPrice      *float64                    `json:"output_price"`
 	CacheWritePrice  *float64                    `json:"cache_write_price"`
 	CacheReadPrice   *float64                    `json:"cache_read_price"`
+	FastMultiplier   *float64                    `json:"fast_multiplier"`
+	FlexMultiplier   *float64                    `json:"flex_multiplier"`
 	ImageInputPrice  *float64                    `json:"image_input_price"`
 	ImageOutputPrice *float64                    `json:"image_output_price"`
 	PerRequestPrice  *float64                    `json:"per_request_price"`
@@ -136,8 +145,9 @@ type channelModelPricingResponse struct {
 }
 
 type channelTimePricingResponse struct {
-	Timezone string                             `json:"timezone"`
-	Periods  []channelTimePricingPeriodResponse `json:"periods"`
+	Timezone     string                             `json:"timezone"`
+	WeekdaysOnly bool                               `json:"weekdays_only"`
+	Periods      []channelTimePricingPeriodResponse `json:"periods"`
 }
 
 type channelTimePricingPeriodResponse struct {
@@ -147,16 +157,20 @@ type channelTimePricingPeriodResponse struct {
 }
 
 type pricingIntervalResponse struct {
-	ID              int64    `json:"id"`
-	MinTokens       int      `json:"min_tokens"`
-	MaxTokens       *int     `json:"max_tokens"`
-	TierLabel       string   `json:"tier_label,omitempty"`
-	InputPrice      *float64 `json:"input_price"`
-	OutputPrice     *float64 `json:"output_price"`
-	CacheWritePrice *float64 `json:"cache_write_price"`
-	CacheReadPrice  *float64 `json:"cache_read_price"`
-	PerRequestPrice *float64 `json:"per_request_price"`
-	SortOrder       int      `json:"sort_order"`
+	ID                   int64    `json:"id"`
+	MinTokens            int      `json:"min_tokens"`
+	MaxTokens            *int     `json:"max_tokens"`
+	TierLabel            string   `json:"tier_label,omitempty"`
+	InputPrice           *float64 `json:"input_price"`
+	OutputPrice          *float64 `json:"output_price"`
+	CacheWritePrice      *float64 `json:"cache_write_price"`
+	CacheReadPrice       *float64 `json:"cache_read_price"`
+	InputMultiplier      *float64 `json:"input_multiplier"`
+	OutputMultiplier     *float64 `json:"output_multiplier"`
+	CacheWriteMultiplier *float64 `json:"cache_write_multiplier"`
+	CacheReadMultiplier  *float64 `json:"cache_read_multiplier"`
+	PerRequestPrice      *float64 `json:"per_request_price"`
+	SortOrder            int      `json:"sort_order"`
 }
 
 type accountStatsPricingRuleResponse struct {
@@ -248,6 +262,8 @@ func pricingToResponse(p *service.ChannelModelPricing) channelModelPricingRespon
 		OutputPrice:      p.OutputPrice,
 		CacheWritePrice:  p.CacheWritePrice,
 		CacheReadPrice:   p.CacheReadPrice,
+		FastMultiplier:   p.FastMultiplier,
+		FlexMultiplier:   p.FlexMultiplier,
 		ImageInputPrice:  p.ImageInputPrice,
 		ImageOutputPrice: p.ImageOutputPrice,
 		PerRequestPrice:  p.PerRequestPrice,
@@ -268,25 +284,33 @@ func timePricingToResponse(value *service.ChannelTimePricing) *channelTimePricin
 			Multiplier: period.Multiplier,
 		})
 	}
-	return &channelTimePricingResponse{Timezone: value.Timezone, Periods: periods}
+	return &channelTimePricingResponse{
+		Timezone:     value.Timezone,
+		WeekdaysOnly: value.WeekdaysOnly,
+		Periods:      periods,
+	}
 }
 
 func intervalToResponse(iv service.PricingInterval) pricingIntervalResponse {
 	return pricingIntervalResponse{
-		ID:              iv.ID,
-		MinTokens:       iv.MinTokens,
-		MaxTokens:       iv.MaxTokens,
-		TierLabel:       iv.TierLabel,
-		InputPrice:      iv.InputPrice,
-		OutputPrice:     iv.OutputPrice,
-		CacheWritePrice: iv.CacheWritePrice,
-		CacheReadPrice:  iv.CacheReadPrice,
-		PerRequestPrice: iv.PerRequestPrice,
-		SortOrder:       iv.SortOrder,
+		ID:                   iv.ID,
+		MinTokens:            iv.MinTokens,
+		MaxTokens:            iv.MaxTokens,
+		TierLabel:            iv.TierLabel,
+		InputPrice:           iv.InputPrice,
+		OutputPrice:          iv.OutputPrice,
+		CacheWritePrice:      iv.CacheWritePrice,
+		CacheReadPrice:       iv.CacheReadPrice,
+		InputMultiplier:      iv.InputMultiplier,
+		OutputMultiplier:     iv.OutputMultiplier,
+		CacheWriteMultiplier: iv.CacheWriteMultiplier,
+		CacheReadMultiplier:  iv.CacheReadMultiplier,
+		PerRequestPrice:      iv.PerRequestPrice,
+		SortOrder:            iv.SortOrder,
 	}
 }
 
-func pricingRequestToService(reqs []channelModelPricingRequest) []service.ChannelModelPricing {
+func pricingRequestToService(reqs []channelModelPricingRequest, allowChannelMultipliers bool) []service.ChannelModelPricing {
 	result := make([]service.ChannelModelPricing, 0, len(reqs))
 	for _, r := range reqs {
 		billingMode := service.BillingMode(r.BillingMode)
@@ -296,18 +320,34 @@ func pricingRequestToService(reqs []channelModelPricingRequest) []service.Channe
 		platform := r.Platform
 		intervals := make([]service.PricingInterval, 0, len(r.Intervals))
 		for _, iv := range r.Intervals {
+			var inputMultiplier, outputMultiplier, cacheWriteMultiplier, cacheReadMultiplier *float64
+			if allowChannelMultipliers {
+				inputMultiplier = iv.InputMultiplier
+				outputMultiplier = iv.OutputMultiplier
+				cacheWriteMultiplier = iv.CacheWriteMultiplier
+				cacheReadMultiplier = iv.CacheReadMultiplier
+			}
 			intervals = append(intervals, service.PricingInterval{
-				MinTokens:       iv.MinTokens,
-				MaxTokens:       iv.MaxTokens,
-				TierLabel:       iv.TierLabel,
-				InputPrice:      iv.InputPrice,
-				OutputPrice:     iv.OutputPrice,
-				CacheWritePrice: iv.CacheWritePrice,
-				CacheReadPrice:  iv.CacheReadPrice,
-				PerRequestPrice: iv.PerRequestPrice,
-				SortOrder:       iv.SortOrder,
+				MinTokens:            iv.MinTokens,
+				MaxTokens:            iv.MaxTokens,
+				TierLabel:            iv.TierLabel,
+				InputPrice:           iv.InputPrice,
+				OutputPrice:          iv.OutputPrice,
+				CacheWritePrice:      iv.CacheWritePrice,
+				CacheReadPrice:       iv.CacheReadPrice,
+				InputMultiplier:      inputMultiplier,
+				OutputMultiplier:     outputMultiplier,
+				CacheWriteMultiplier: cacheWriteMultiplier,
+				CacheReadMultiplier:  cacheReadMultiplier,
+				PerRequestPrice:      iv.PerRequestPrice,
+				SortOrder:            iv.SortOrder,
 			})
 		}
+		var fastMultiplier, flexMultiplier *float64
+		if allowChannelMultipliers {
+			fastMultiplier = r.FastMultiplier
+			flexMultiplier = r.FlexMultiplier
+		}
 		result = append(result, service.ChannelModelPricing{
 			Platform:         platform,
 			Models:           r.Models,
@@ -316,6 +356,8 @@ func pricingRequestToService(reqs []channelModelPricingRequest) []service.Channe
 			OutputPrice:      r.OutputPrice,
 			CacheWritePrice:  r.CacheWritePrice,
 			CacheReadPrice:   r.CacheReadPrice,
+			FastMultiplier:   fastMultiplier,
+			FlexMultiplier:   flexMultiplier,
 			ImageInputPrice:  r.ImageInputPrice,
 			ImageOutputPrice: r.ImageOutputPrice,
 			PerRequestPrice:  r.PerRequestPrice,
@@ -338,7 +380,11 @@ func timePricingRequestToService(value *channelTimePricingRequest) *service.Chan
 			Multiplier: period.Multiplier,
 		})
 	}
-	return &service.ChannelTimePricing{Timezone: value.Timezone, Periods: periods}
+	return &service.ChannelTimePricing{
+		Timezone:     value.Timezone,
+		WeekdaysOnly: value.WeekdaysOnly,
+		Periods:      periods,
+	}
 }
 
 func accountStatsPricingRuleRequestToService(r accountStatsPricingRuleRequest) service.AccountStatsPricingRule {
@@ -346,7 +392,7 @@ func accountStatsPricingRuleRequestToService(r accountStatsPricingRuleRequest) s
 		Name:       r.Name,
 		GroupIDs:   r.GroupIDs,
 		AccountIDs: r.AccountIDs,
-		Pricing:    pricingRequestToService(r.Pricing),
+		Pricing:    pricingRequestToService(r.Pricing, false),
 	}
 }
 
@@ -407,7 +453,7 @@ func (h *ChannelHandler) Create(c *gin.Context) {
 		return
 	}
 
-	pricing := pricingRequestToService(req.ModelPricing)
+	pricing := pricingRequestToService(req.ModelPricing, true)
 	// Main model_pricing requires a platform; default to anthropic for backward compatibility.
 	for i := range pricing {
 		if pricing[i].Platform == "" {
@@ -481,7 +527,7 @@ func (h *ChannelHandler) Update(c *gin.Context) {
 		ApplyPricingToAccountStats: req.ApplyPricingToAccountStats,
 	}
 	if req.ModelPricing != nil {
-		pricing := pricingRequestToService(*req.ModelPricing)
+		pricing := pricingRequestToService(*req.ModelPricing, true)
 		for i := range pricing {
 			if pricing[i].Platform == "" {
 				pricing[i].Platform = service.PlatformAnthropic

backend/internal/handler/admin/channel_handler_test.go

@@ -305,7 +305,7 @@ func TestPricingRequestToService_Defaults(t *testing.T) {
 
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
-			result := pricingRequestToService([]channelModelPricingRequest{tt.req})
+			result := pricingRequestToService([]channelModelPricingRequest{tt.req}, true)
 			require.Len(t, result, 1)
 			switch tt.wantField {
 			case "BillingMode":
@@ -332,7 +332,7 @@ func TestPricingRequestToService_WithAllFields(t *testing.T) {
 		},
 	}
 
-	result := pricingRequestToService(reqs)
+	result := pricingRequestToService(reqs, true)
 	require.Len(t, result, 1)
 	r := result[0]
 	require.Equal(t, "openai", r.Platform)
@@ -373,7 +373,7 @@ func TestPricingRequestToService_WithIntervals(t *testing.T) {
 		},
 	}
 
-	result := pricingRequestToService(reqs)
+	result := pricingRequestToService(reqs, true)
 	require.Len(t, result, 1)
 	require.Len(t, result[0].Intervals, 2)
 
@@ -396,7 +396,7 @@ func TestPricingRequestToService_WithIntervals(t *testing.T) {
 }
 
 func TestPricingRequestToService_EmptySlice(t *testing.T) {
-	result := pricingRequestToService([]channelModelPricingRequest{})
+	result := pricingRequestToService([]channelModelPricingRequest{}, true)
 	require.NotNil(t, result)
 	require.Empty(t, result)
 }
@@ -410,7 +410,7 @@ func TestPricingRequestToService_NilPriceFields(t *testing.T) {
 		},
 	}
 
-	result := pricingRequestToService(reqs)
+	result := pricingRequestToService(reqs, true)
 	require.Len(t, result, 1)
 	r := result[0]
 	require.Nil(t, r.InputPrice)
@@ -426,28 +426,67 @@ func TestPricingRequestToService_TimePricing(t *testing.T) {
 		Models:      []string{"gpt-5"},
 		BillingMode: "token",
 		TimePricing: &channelTimePricingRequest{
-			Timezone: "Asia/Shanghai",
+			Timezone:     "Asia/Shanghai",
+			WeekdaysOnly: true,
 			Periods: []channelTimePricingPeriodRequest{{
 				StartTime: "09:00", EndTime: "12:00", Multiplier: 2,
 			}},
 		},
 	}
 
-	got := pricingRequestToService([]channelModelPricingRequest{req})
+	got := pricingRequestToService([]channelModelPricingRequest{req}, true)
 	require.Equal(t, "Asia/Shanghai", got[0].TimePricing.Timezone)
+	require.True(t, got[0].TimePricing.WeekdaysOnly)
 	require.Equal(t, 2.0, got[0].TimePricing.Periods[0].Multiplier)
 }
 
 func TestPricingRequestToService_TimePricingNil(t *testing.T) {
-	got := pricingRequestToService([]channelModelPricingRequest{{Models: []string{"gpt-5"}}})
+	got := pricingRequestToService([]channelModelPricingRequest{{Models: []string{"gpt-5"}}}, true)
 	require.Nil(t, got[0].TimePricing)
 }
 
+// 账号成本统计规则不支持倍率：allowChannelMultipliers=false 时必须丢弃，
+// 避免渠道倍率意外污染账号成本口径。
+func TestPricingRequestToService_MultipliersGatedByFlag(t *testing.T) {
+	req := channelModelPricingRequest{
+		Models:         []string{"gpt-5"},
+		BillingMode:    "token",
+		FastMultiplier: float64Ptr(2.5),
+		FlexMultiplier: float64Ptr(0.5),
+		Intervals: []pricingIntervalRequest{{
+			MinTokens:            272000,
+			InputMultiplier:      float64Ptr(2),
+			OutputMultiplier:     float64Ptr(1.5),
+			CacheWriteMultiplier: float64Ptr(2),
+			CacheReadMultiplier:  float64Ptr(2),
+		}},
+	}
+
+	allowed := pricingRequestToService([]channelModelPricingRequest{req}, true)
+	require.Equal(t, float64Ptr(2.5), allowed[0].FastMultiplier)
+	require.Equal(t, float64Ptr(0.5), allowed[0].FlexMultiplier)
+	require.Equal(t, float64Ptr(2), allowed[0].Intervals[0].InputMultiplier)
+	require.Equal(t, float64Ptr(1.5), allowed[0].Intervals[0].OutputMultiplier)
+	require.Equal(t, float64Ptr(2), allowed[0].Intervals[0].CacheWriteMultiplier)
+	require.Equal(t, float64Ptr(2), allowed[0].Intervals[0].CacheReadMultiplier)
+
+	dropped := pricingRequestToService([]channelModelPricingRequest{req}, false)
+	require.Nil(t, dropped[0].FastMultiplier)
+	require.Nil(t, dropped[0].FlexMultiplier)
+	require.Nil(t, dropped[0].Intervals[0].InputMultiplier)
+	require.Nil(t, dropped[0].Intervals[0].OutputMultiplier)
+	require.Nil(t, dropped[0].Intervals[0].CacheWriteMultiplier)
+	require.Nil(t, dropped[0].Intervals[0].CacheReadMultiplier)
+	// 非倍率字段不受开关影响
+	require.Equal(t, 272000, dropped[0].Intervals[0].MinTokens)
+}
+
 func TestPricingToResponse_TimePricing(t *testing.T) {
 	got := pricingToResponse(&service.ChannelModelPricing{
 		BillingMode: service.BillingModeToken,
 		TimePricing: &service.ChannelTimePricing{
-			Timezone: "Asia/Shanghai",
+			Timezone:     "Asia/Shanghai",
+			WeekdaysOnly: true,
 			Periods: []service.ChannelTimePricingPeriod{{
 				StartTime: "14:00", EndTime: "18:00", Multiplier: 1.25,
 			}},
@@ -456,6 +495,7 @@ func TestPricingToResponse_TimePricing(t *testing.T) {
 
 	require.NotNil(t, got.TimePricing)
 	require.Equal(t, "Asia/Shanghai", got.TimePricing.Timezone)
+	require.True(t, got.TimePricing.WeekdaysOnly)
 	require.Equal(t, 1.25, got.TimePricing.Periods[0].Multiplier)
 }
 

backend/internal/handler/admin/dashboard_handler.go

@@ -262,13 +262,18 @@ func (h *DashboardHandler) GetUsageTrend(c *gin.Context) {
 			return
 		}
 	}
-	upstreamModelMismatch, err := parseOptionalBoolDashboardFilter(c, "upstream_model_mismatch")
+	nativeCompactionV2, err := parseOptionalBoolDashboardFilter(c, "native_compaction_v2")
+	if err != nil {
+		response.BadRequest(c, "Invalid native_compaction_v2 value, use true or false")
+		return
+	}
+	upstreamModelMismatch, err = parseOptionalBoolDashboardFilter(c, "upstream_model_mismatch")
 	if err != nil {
 		response.BadRequest(c, "Invalid upstream_model_mismatch value, use true or false")
 		return
 	}
 
-	trend, hit, err := h.getUsageTrendCached(c.Request.Context(), startTime, endTime, granularity, userID, apiKeyID, accountID, groupID, model, requestType, stream, billingType, upstreamModelMismatch)
+	trend, hit, err := h.getUsageTrendCached(c.Request.Context(), startTime, endTime, granularity, userID, apiKeyID, accountID, groupID, model, requestType, stream, nativeCompactionV2, billingType, upstreamModelMismatch)
 	if err != nil {
 		response.Error(c, 500, "Failed to get usage trend")
 		return
@@ -349,13 +354,18 @@ func (h *DashboardHandler) GetModelStats(c *gin.Context) {
 			return
 		}
 	}
-	upstreamModelMismatch, err := parseOptionalBoolDashboardFilter(c, "upstream_model_mismatch")
+	nativeCompactionV2, err := parseOptionalBoolDashboardFilter(c, "native_compaction_v2")
+	if err != nil {
+		response.BadRequest(c, "Invalid native_compaction_v2 value, use true or false")
+		return
+	}
+	upstreamModelMismatch, err = parseOptionalBoolDashboardFilter(c, "upstream_model_mismatch")
 	if err != nil {
 		response.BadRequest(c, "Invalid upstream_model_mismatch value, use true or false")
 		return
 	}
 
-	stats, hit, err := h.getModelStatsCached(c.Request.Context(), startTime, endTime, userID, apiKeyID, accountID, groupID, modelSource, requestType, stream, billingType, upstreamModelMismatch)
+	stats, hit, err := h.getModelStatsCached(c.Request.Context(), startTime, endTime, userID, apiKeyID, accountID, groupID, modelSource, requestType, stream, nativeCompactionV2, billingType, upstreamModelMismatch)
 	if err != nil {
 		response.Error(c, 500, "Failed to get model statistics")
 		return
@@ -426,13 +436,18 @@ func (h *DashboardHandler) GetGroupStats(c *gin.Context) {
 			return
 		}
 	}
-	upstreamModelMismatch, err := parseOptionalBoolDashboardFilter(c, "upstream_model_mismatch")
+	nativeCompactionV2, err := parseOptionalBoolDashboardFilter(c, "native_compaction_v2")
+	if err != nil {
+		response.BadRequest(c, "Invalid native_compaction_v2 value, use true or false")
+		return
+	}
+	upstreamModelMismatch, err = parseOptionalBoolDashboardFilter(c, "upstream_model_mismatch")
 	if err != nil {
 		response.BadRequest(c, "Invalid upstream_model_mismatch value, use true or false")
 		return
 	}
 
-	stats, hit, err := h.getGroupStatsCached(c.Request.Context(), startTime, endTime, userID, apiKeyID, accountID, groupID, requestType, stream, billingType, upstreamModelMismatch)
+	stats, hit, err := h.getGroupStatsCached(c.Request.Context(), startTime, endTime, userID, apiKeyID, accountID, groupID, requestType, stream, nativeCompactionV2, billingType, upstreamModelMismatch)
 	if err != nil {
 		response.Error(c, 500, "Failed to get group statistics")
 		return
@@ -701,6 +716,14 @@ func (h *DashboardHandler) GetUserBreakdown(c *gin.Context) {
 			dim.Stream = &s
 		}
 	}
+	if v := strings.TrimSpace(c.Query("native_compaction_v2")); v != "" {
+		value, err := strconv.ParseBool(v)
+		if err != nil {
+			response.BadRequest(c, "Invalid native_compaction_v2 value, use true or false")
+			return
+		}
+		dim.NativeCompactionV2 = &value
+	}
 	if v := c.Query("billing_type"); v != "" {
 		if bt, err := strconv.ParseInt(v, 10, 8); err == nil {
 			btVal := int8(bt)

backend/internal/handler/admin/dashboard_handler_request_type_test.go

@@ -15,16 +15,19 @@ import (
 
 type dashboardUsageRepoCapture struct {
 	service.UsageLogRepository
-	trendRequestType *int16
-	trendStream      *bool
-	modelRequestType *int16
-	modelStream      *bool
-	trendMismatch    *bool
-	modelMismatch    *bool
-	groupMismatch    *bool
-	rankingLimit     int
-	ranking          []usagestats.UserSpendingRankingItem
-	rankingTotal     float64
+	trendRequestType      *int16
+	trendStream           *bool
+	trendNativeCompaction *bool
+	modelRequestType      *int16
+	modelStream           *bool
+	modelNativeCompaction *bool
+	groupNativeCompaction *bool
+	trendMismatch         *bool
+	modelMismatch         *bool
+	groupMismatch         *bool
+	rankingLimit          int
+	ranking               []usagestats.UserSpendingRankingItem
+	rankingTotal          float64
 }
 
 func (s *dashboardUsageRepoCapture) GetUsageTrendWithUsageFilters(
@@ -35,6 +38,7 @@ func (s *dashboardUsageRepoCapture) GetUsageTrendWithUsageFilters(
 ) ([]usagestats.TrendDataPoint, error) {
 	s.trendRequestType = filters.RequestType
 	s.trendStream = filters.Stream
+	s.trendNativeCompaction = filters.NativeCompactionV2
 	s.trendMismatch = filters.UpstreamModelMismatch
 	return []usagestats.TrendDataPoint{}, nil
 }
@@ -62,6 +66,7 @@ func (s *dashboardUsageRepoCapture) GetModelStatsWithUsageFiltersBySource(
 ) ([]usagestats.ModelStat, error) {
 	s.modelRequestType = filters.RequestType
 	s.modelStream = filters.Stream
+	s.modelNativeCompaction = filters.NativeCompactionV2
 	s.modelMismatch = filters.UpstreamModelMismatch
 	return []usagestats.ModelStat{}, nil
 }
@@ -71,6 +76,7 @@ func (s *dashboardUsageRepoCapture) GetGroupStatsWithUsageFilters(
 	startTime, endTime time.Time,
 	filters usagestats.UsageLogFilters,
 ) ([]usagestats.GroupStat, error) {
+	s.groupNativeCompaction = filters.NativeCompactionV2
 	s.groupMismatch = filters.UpstreamModelMismatch
 	return []usagestats.GroupStat{}, nil
 }
@@ -208,6 +214,48 @@ func TestDashboardModelStatsValidModelSource(t *testing.T) {
 	require.Equal(t, http.StatusOK, rec.Code)
 }
 
+func TestDashboardNativeCompactionFilterPropagatesAlongsideTransport(t *testing.T) {
+	resetDashboardReadCachesForTest()
+	repo := &dashboardUsageRepoCapture{}
+	router := newDashboardRequestTypeTestRouter(repo)
+
+	for _, path := range []string{
+		"/admin/dashboard/trend?request_type=stream&native_compaction_v2=true",
+		"/admin/dashboard/models?request_type=stream&native_compaction_v2=true",
+		"/admin/dashboard/groups?request_type=stream&native_compaction_v2=true",
+	} {
+		req := httptest.NewRequest(http.MethodGet, path, nil)
+		rec := httptest.NewRecorder()
+		router.ServeHTTP(rec, req)
+		require.Equal(t, http.StatusOK, rec.Code, path)
+	}
+
+	require.NotNil(t, repo.trendNativeCompaction)
+	require.True(t, *repo.trendNativeCompaction)
+	require.NotNil(t, repo.modelNativeCompaction)
+	require.True(t, *repo.modelNativeCompaction)
+	require.NotNil(t, repo.groupNativeCompaction)
+	require.True(t, *repo.groupNativeCompaction)
+	require.NotNil(t, repo.trendRequestType)
+	require.Equal(t, int16(service.RequestTypeStream), *repo.trendRequestType)
+}
+
+func TestDashboardNativeCompactionFilterRejectsInvalidBoolean(t *testing.T) {
+	repo := &dashboardUsageRepoCapture{}
+	router := newDashboardRequestTypeTestRouter(repo)
+
+	for _, path := range []string{
+		"/admin/dashboard/trend?native_compaction_v2=invalid",
+		"/admin/dashboard/models?native_compaction_v2=invalid",
+		"/admin/dashboard/groups?native_compaction_v2=invalid",
+	} {
+		req := httptest.NewRequest(http.MethodGet, path, nil)
+		rec := httptest.NewRecorder()
+		router.ServeHTTP(rec, req)
+		require.Equal(t, http.StatusBadRequest, rec.Code, path)
+	}
+}
+
 func TestDashboardModelAuditFilterPropagatesToTrendModelAndGroupQueries(t *testing.T) {
 	resetDashboardReadCachesForTest()
 	repo := &dashboardUsageRepoCapture{}

backend/internal/handler/admin/dashboard_query_cache.go

@@ -28,6 +28,7 @@ type dashboardTrendCacheKey struct {
 	Model                 string `json:"model"`
 	RequestType           *int16 `json:"request_type"`
 	Stream                *bool  `json:"stream"`
+	NativeCompactionV2    *bool  `json:"native_compaction_v2"`
 	BillingType           *int8  `json:"billing_type"`
 	UpstreamModelMismatch *bool  `json:"upstream_model_mismatch"`
 }
@@ -42,6 +43,7 @@ type dashboardModelGroupCacheKey struct {
 	ModelSource           string `json:"model_source,omitempty"`
 	RequestType           *int16 `json:"request_type"`
 	Stream                *bool  `json:"stream"`
+	NativeCompactionV2    *bool  `json:"native_compaction_v2"`
 	BillingType           *int8  `json:"billing_type"`
 	UpstreamModelMismatch *bool  `json:"upstream_model_mismatch"`
 }
@@ -85,6 +87,7 @@ func (h *DashboardHandler) getUsageTrendCached(
 	model string,
 	requestType *int16,
 	stream *bool,
+	nativeCompactionV2 *bool,
 	billingType *int8,
 	upstreamModelMismatch *bool,
 ) ([]usagestats.TrendDataPoint, bool, error) {
@@ -99,13 +102,14 @@ func (h *DashboardHandler) getUsageTrendCached(
 		Model:                 model,
 		RequestType:           requestType,
 		Stream:                stream,
+		NativeCompactionV2:    nativeCompactionV2,
 		BillingType:           billingType,
 		UpstreamModelMismatch: upstreamModelMismatch,
 	})
 	entry, hit, err := dashboardTrendCache.GetOrLoad(key, func() (any, error) {
 		return h.dashboardService.GetUsageTrendWithUsageFilters(ctx, startTime, endTime, granularity, usagestats.UsageLogFilters{
 			UserID: userID, APIKeyID: apiKeyID, AccountID: accountID, GroupID: groupID,
-			Model: model, RequestType: requestType, Stream: stream, BillingType: billingType,
+			Model: model, RequestType: requestType, Stream: stream, NativeCompactionV2: nativeCompactionV2, BillingType: billingType,
 			UpstreamModelMismatch: upstreamModelMismatch,
 		})
 	})
@@ -123,6 +127,7 @@ func (h *DashboardHandler) getModelStatsCached(
 	modelSource string,
 	requestType *int16,
 	stream *bool,
+	nativeCompactionV2 *bool,
 	billingType *int8,
 	upstreamModelMismatch *bool,
 ) ([]usagestats.ModelStat, bool, error) {
@@ -136,13 +141,14 @@ func (h *DashboardHandler) getModelStatsCached(
 		ModelSource:           usagestats.NormalizeModelSource(modelSource),
 		RequestType:           requestType,
 		Stream:                stream,
+		NativeCompactionV2:    nativeCompactionV2,
 		BillingType:           billingType,
 		UpstreamModelMismatch: upstreamModelMismatch,
 	})
 	entry, hit, err := dashboardModelStatsCache.GetOrLoad(key, func() (any, error) {
 		return h.dashboardService.GetModelStatsWithUsageFiltersBySource(ctx, startTime, endTime, usagestats.UsageLogFilters{
 			UserID: userID, APIKeyID: apiKeyID, AccountID: accountID, GroupID: groupID,
-			RequestType: requestType, Stream: stream, BillingType: billingType,
+			RequestType: requestType, Stream: stream, NativeCompactionV2: nativeCompactionV2, BillingType: billingType,
 			UpstreamModelMismatch: upstreamModelMismatch,
 		}, modelSource)
 	})
@@ -159,6 +165,7 @@ func (h *DashboardHandler) getGroupStatsCached(
 	userID, apiKeyID, accountID, groupID int64,
 	requestType *int16,
 	stream *bool,
+	nativeCompactionV2 *bool,
 	billingType *int8,
 	upstreamModelMismatch *bool,
 ) ([]usagestats.GroupStat, bool, error) {
@@ -171,13 +178,14 @@ func (h *DashboardHandler) getGroupStatsCached(
 		GroupID:               groupID,
 		RequestType:           requestType,
 		Stream:                stream,
+		NativeCompactionV2:    nativeCompactionV2,
 		BillingType:           billingType,
 		UpstreamModelMismatch: upstreamModelMismatch,
 	})
 	entry, hit, err := dashboardGroupStatsCache.GetOrLoad(key, func() (any, error) {
 		return h.dashboardService.GetGroupStatsWithUsageFilters(ctx, startTime, endTime, usagestats.UsageLogFilters{
 			UserID: userID, APIKeyID: apiKeyID, AccountID: accountID, GroupID: groupID,
-			RequestType: requestType, Stream: stream, BillingType: billingType,
+			RequestType: requestType, Stream: stream, NativeCompactionV2: nativeCompactionV2, BillingType: billingType,
 			UpstreamModelMismatch: upstreamModelMismatch,
 		})
 	})

backend/internal/handler/admin/dashboard_snapshot_v2_handler.go

@@ -44,6 +44,7 @@ type dashboardSnapshotV2Filters struct {
 	Model                 string
 	RequestType           *int16
 	Stream                *bool
+	NativeCompactionV2    *bool
 	BillingType           *int8
 	UpstreamModelMismatch *bool
 }
@@ -59,6 +60,7 @@ type dashboardSnapshotV2CacheKey struct {
 	Model                 string `json:"model"`
 	RequestType           *int16 `json:"request_type"`
 	Stream                *bool  `json:"stream"`
+	NativeCompactionV2    *bool  `json:"native_compaction_v2"`
 	BillingType           *int8  `json:"billing_type"`
 	UpstreamModelMismatch *bool  `json:"upstream_model_mismatch"`
 	IncludeStats          bool   `json:"include_stats"`
@@ -105,6 +107,7 @@ func (h *DashboardHandler) GetSnapshotV2(c *gin.Context) {
 		Model:                 filters.Model,
 		RequestType:           filters.RequestType,
 		Stream:                filters.Stream,
+		NativeCompactionV2:    filters.NativeCompactionV2,
 		BillingType:           filters.BillingType,
 		UpstreamModelMismatch: filters.UpstreamModelMismatch,
 		IncludeStats:          includeStats,
@@ -186,6 +189,7 @@ func (h *DashboardHandler) buildSnapshotV2Response(
 			filters.Model,
 			filters.RequestType,
 			filters.Stream,
+			filters.NativeCompactionV2,
 			filters.BillingType,
 			filters.UpstreamModelMismatch,
 		)
@@ -207,6 +211,7 @@ func (h *DashboardHandler) buildSnapshotV2Response(
 			usagestats.ModelSourceRequested,
 			filters.RequestType,
 			filters.Stream,
+			filters.NativeCompactionV2,
 			filters.BillingType,
 			filters.UpstreamModelMismatch,
 		)
@@ -227,6 +232,7 @@ func (h *DashboardHandler) buildSnapshotV2Response(
 			filters.GroupID,
 			filters.RequestType,
 			filters.Stream,
+			filters.NativeCompactionV2,
 			filters.BillingType,
 			filters.UpstreamModelMismatch,
 		)
@@ -296,6 +302,14 @@ func parseDashboardSnapshotV2Filters(c *gin.Context) (*dashboardSnapshotV2Filter
 		filters.Stream = &streamVal
 	}
 
+	if nativeCompactionV2Str := strings.TrimSpace(c.Query("native_compaction_v2")); nativeCompactionV2Str != "" {
+		value, err := strconv.ParseBool(nativeCompactionV2Str)
+		if err != nil {
+			return nil, err
+		}
+		filters.NativeCompactionV2 = &value
+	}
+
 	if billingTypeStr := strings.TrimSpace(c.Query("billing_type")); billingTypeStr != "" {
 		v, err := strconv.ParseInt(billingTypeStr, 10, 8)
 		if err != nil {

backend/internal/handler/admin/grok_oauth_handler.go

@@ -615,8 +615,10 @@ func (h *GrokOAuthHandler) ResetQuota(c *gin.Context) {
 		response.BadRequest(c, "grok quota service is not enabled")
 		return
 	}
+	// ResetQuota 恒返回 GROK_QUOTA_RESET_UNSUPPORTED（xAI 无 OAuth 配额重置接口），err != nil 恒真为预期。
+	//nolint:staticcheck // SA4023
 	result, err := h.quotaService.ResetQuota(c.Request.Context(), accountID)
-	if err != nil {
+	if err != nil { //nolint:staticcheck // SA4023
 		response.ErrorFrom(c, err)
 		return
 	}

backend/internal/handler/admin/group_handler.go

@@ -81,8 +81,8 @@ func (f optionalLimitField) ToServiceInput() *float64 {
 	if f.value != nil {
 		return f.value
 	}
-	zero := 0.0
-	return &zero
+	unlimited := -1.0
+	return &unlimited
 }
 
 // NewGroupHandler creates a new admin group handler
@@ -234,7 +234,7 @@ type UpdateGroupRequest struct {
 type CompositeRouteRequest struct {
 	PublicModel    string `json:"public_model" binding:"required"`
 	MatchType      string `json:"match_type" binding:"omitempty,oneof=exact prefix"`
-	TargetPlatform string `json:"target_platform" binding:"required,oneof=anthropic openai gemini antigravity grok"`
+	TargetPlatform string `json:"target_platform" binding:"required,oneof=anthropic openai gemini antigravity grok kimi zhipu deepseek"`
 	UpstreamModel  string `json:"upstream_model"`
 	Endpoint       string `json:"endpoint" binding:"omitempty,oneof=any messages count_tokens responses chat_completions embeddings images gemini"`
 	Priority       int    `json:"priority"`

backend/internal/handler/admin/group_handler_limit_test.go

@@ -0,0 +1,33 @@
+package admin
+
+import (
+	"encoding/json"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+)
+
+func TestUpdateGroupRequestLimitFieldsTriState(t *testing.T) {
+	t.Run("omitted means unchanged", func(t *testing.T) {
+		var req UpdateGroupRequest
+		require.NoError(t, json.Unmarshal([]byte(`{}`), &req))
+		require.Nil(t, req.DailyLimitUSD.ToServiceInput())
+		require.Nil(t, req.WeeklyLimitUSD.ToServiceInput())
+		require.Nil(t, req.MonthlyLimitUSD.ToServiceInput())
+	})
+
+	t.Run("null means unlimited", func(t *testing.T) {
+		var req UpdateGroupRequest
+		require.NoError(t, json.Unmarshal([]byte(`{"daily_limit_usd":null}`), &req))
+		limit := req.DailyLimitUSD.ToServiceInput()
+		require.NotNil(t, limit)
+		require.Negative(t, *limit)
+	})
+
+	t.Run("number is preserved", func(t *testing.T) {
+		var req UpdateGroupRequest
+		require.NoError(t, json.Unmarshal([]byte(`{"weekly_limit_usd":0,"monthly_limit_usd":42.5}`), &req))
+		require.Equal(t, 0.0, *req.WeeklyLimitUSD.ToServiceInput())
+		require.Equal(t, 42.5, *req.MonthlyLimitUSD.ToServiceInput())
+	})
+}

backend/internal/handler/admin/group_handler_platform_test.go

@@ -71,13 +71,11 @@ func TestGroupPlatformBinding_RejectsInvalidPlatforms(t *testing.T) {
 	}
 }
 
-// 守住 composite 路由目标不放行 CN:CN 平台不可作为 composite 路由目标
-// （DetectModelPlatform/isConcreteRequestPlatform 均无 CN 分支,放行即打开半实现路径）。
-func TestCompositeRouteTargetPlatform_StillExcludesCNProviders(t *testing.T) {
+func TestCompositeRouteTargetPlatform_AllowsCNProviders(t *testing.T) {
 	for _, platform := range []string{"kimi", "zhipu", "deepseek"} {
 		var req CompositeRouteRequest
 		body := fmt.Sprintf(`{"public_model":"m","target_platform":%q}`, platform)
-		require.Error(t, bindGroupPlatformJSON(t, &req, body),
-			"composite target_platform %q 应保持被拒", platform)
+		require.NoError(t, bindGroupPlatformJSON(t, &req, body))
+		require.Equal(t, platform, req.TargetPlatform)
 	}
 }

backend/internal/handler/admin/openai_oauth_handler.go

@@ -27,6 +27,7 @@ type OpenAIOAuthHandler struct {
 type openAIQuotaService interface {
 	QueryUsage(ctx context.Context, accountID int64) (*service.OpenAIQuotaUsage, error)
 	CacheResetCreditsSnapshot(ctx context.Context, accountID int64, credits *service.OpenAIRateLimitResetCredits) error
+	CachePostResetSnapshot(ctx context.Context, accountID int64, usage *service.OpenAIQuotaUsage) error
 	ResetCredit(ctx context.Context, accountID int64) (*service.OpenAIQuotaResetResult, error)
 }
 
@@ -34,12 +35,6 @@ type openAIAccountStateRecoverer interface {
 	RecoverAccountState(ctx context.Context, accountID int64, options service.AccountRecoveryOptions) (*service.SuccessfulTestRecoveryResult, error)
 }
 
-const (
-	openAIQuotaResetWarningCacheRefreshFailed    = "reset_credit_cache_refresh_failed"
-	openAIQuotaResetWarningAccountRecoveryFailed = "account_state_recovery_failed"
-	openAIQuotaResetWarningAccountRefreshFailed  = "account_state_refresh_failed"
-)
-
 // openAIQuotaResetPostProcessTimeout bounds the work performed AFTER the
 // (non-refundable) reset credit has already been consumed upstream. The whole
 // request must stay comfortably inside the panel HTTP client timeout, otherwise
@@ -493,6 +488,7 @@ func (h *OpenAIOAuthHandler) QueryQuota(c *gin.Context) {
 		response.ErrorFrom(c, err)
 		return
 	}
+	service.NotifyOpenAIAutoResetCredit(accountID)
 	response.Success(c, usage)
 }
 
@@ -523,6 +519,7 @@ func (h *OpenAIOAuthHandler) RefreshQuota(c *gin.Context) {
 		response.Error(c, http.StatusInternalServerError, "openai quota query returned an empty result")
 		return
 	}
+	service.NotifyOpenAIAutoResetCredit(accountID)
 
 	refreshResponse := openAIQuotaRefreshResponse{OpenAIQuotaUsage: *usage}
 	// A failed snapshot write leaves the previous cache intact — report it as a
@@ -600,54 +597,19 @@ func (h *OpenAIOAuthHandler) ResetQuota(c *gin.Context) {
 	postCtx, cancelPost := openAIQuotaResetPostProcessContext(c.Request.Context())
 	defer cancelPost()
 
-	// Step 1 — unblocking the account is the whole point of consuming a credit
-	// (#3672 / #3740), so it runs FIRST and is never gated on the display cache.
-	// Recovery is DB-only and leaves the manual `schedulable` switch untouched.
-	if h.rateLimitService == nil {
-		resetResponse.WarningCode = openAIQuotaResetWarningAccountRecoveryFailed
-		response.Success(c, resetResponse)
-		return
+	postResult := service.RunOpenAIQuotaResetPostProcess(
+		postCtx,
+		accountID,
+		h.quotaService,
+		h.rateLimitService,
+		h.adminService.GetAccount,
+	)
+	resetResponse.Quota = postResult.Quota
+	resetResponse.CacheRefreshed = postResult.CacheRefreshed
+	resetResponse.AccountStateRecovered = postResult.AccountStateRecovered
+	resetResponse.WarningCode = postResult.WarningCode
+	if postResult.Account != nil {
+		resetResponse.Account = dto.AccountFromService(postResult.Account)
 	}
-	if _, err := h.rateLimitService.RecoverAccountState(postCtx, accountID, service.AccountRecoveryOptions{
-		InvalidateToken: true,
-	}); err != nil {
-		// Recovery failures are almost always storage-level; the remaining steps
-		// share that dependency, so stop here instead of compounding the failure.
-		slog.Warn("openai_quota_reset_account_recovery_failed", "account_id", accountID, "error", err)
-		resetResponse.WarningCode = openAIQuotaResetWarningAccountRecoveryFailed
-		response.Success(c, resetResponse)
-		return
-	}
-	resetResponse.AccountStateRecovered = true
-
-	// Step 2 — refresh the reset-credit display cache. A failure here is reported
-	// but must not hide the recovered account row produced by step 3.
-	usage, usageErr := h.quotaService.QueryUsage(postCtx, accountID)
-	switch {
-	case usageErr != nil || usage == nil:
-		slog.Warn("openai_quota_reset_cache_refresh_failed", "account_id", accountID, "error", usageErr)
-		resetResponse.WarningCode = openAIQuotaResetWarningCacheRefreshFailed
-	default:
-		if err := h.quotaService.CacheResetCreditsSnapshot(postCtx, accountID, usage.RateLimitResetCredits); err != nil {
-			slog.Warn("openai_quota_reset_cache_refresh_failed", "account_id", accountID, "error", err)
-			resetResponse.WarningCode = openAIQuotaResetWarningCacheRefreshFailed
-		} else {
-			resetResponse.Quota = usage
-			resetResponse.CacheRefreshed = true
-		}
-	}
-
-	// Step 3 — hand back the post-recovery account row so the list drops the
-	// stale rate-limit badge without waiting for the next poll.
-	account, err := h.adminService.GetAccount(postCtx, accountID)
-	if err != nil {
-		slog.Warn("openai_quota_reset_account_refresh_failed", "account_id", accountID, "error", err)
-		if resetResponse.WarningCode == "" {
-			resetResponse.WarningCode = openAIQuotaResetWarningAccountRefreshFailed
-		}
-		response.Success(c, resetResponse)
-		return
-	}
-	resetResponse.Account = dto.AccountFromService(account)
 	response.Success(c, resetResponse)
 }

backend/internal/handler/admin/openai_oauth_handler_reset_quota_test.go

@@ -48,6 +48,12 @@ func (s *openAIQuotaWorkflowStub) CacheResetCreditsSnapshot(ctx context.Context,
 	return s.cacheErr
 }
 
+func (s *openAIQuotaWorkflowStub) CachePostResetSnapshot(ctx context.Context, _ int64, _ *service.OpenAIQuotaUsage) error {
+	s.cacheCalls++
+	s.cacheCtxErr = ctx.Err()
+	return s.cacheErr
+}
+
 type openAIAccountStateRecovererStub struct {
 	err         error
 	calls       int
@@ -215,7 +221,7 @@ func TestOpenAIResetQuota_RecoveryFailureStopsWorkflow(t *testing.T) {
 	status, envelope := performOpenAIQuotaResetRequest(t, handler)
 
 	require.Equal(t, http.StatusOK, status)
-	require.Equal(t, openAIQuotaResetWarningAccountRecoveryFailed, envelope.Data.WarningCode)
+	require.Equal(t, service.OpenAIQuotaResetWarningAccountRecoveryFailed, envelope.Data.WarningCode)
 	require.False(t, envelope.Data.AccountStateRecovered)
 	require.False(t, envelope.Data.CacheRefreshed)
 	require.Nil(t, envelope.Data.Quota)
@@ -237,7 +243,7 @@ func TestOpenAIResetQuota_MissingRecovererReportsRecoveryFailure(t *testing.T) {
 	status, envelope := performOpenAIQuotaResetRequest(t, handler)
 
 	require.Equal(t, http.StatusOK, status)
-	require.Equal(t, openAIQuotaResetWarningAccountRecoveryFailed, envelope.Data.WarningCode)
+	require.Equal(t, service.OpenAIQuotaResetWarningAccountRecoveryFailed, envelope.Data.WarningCode)
 	require.False(t, envelope.Data.AccountStateRecovered)
 	require.Zero(t, quota.queryCalls)
 	require.Zero(t, adminService.calls)
@@ -260,7 +266,7 @@ func TestOpenAIResetQuota_QueryFailureStillRecoversAndReturnsAccount(t *testing.
 	status, envelope := performOpenAIQuotaResetRequest(t, handler)
 
 	require.Equal(t, http.StatusOK, status)
-	require.Equal(t, openAIQuotaResetWarningCacheRefreshFailed, envelope.Data.WarningCode)
+	require.Equal(t, service.OpenAIQuotaResetWarningCacheRefreshFailed, envelope.Data.WarningCode)
 	require.True(t, envelope.Data.AccountStateRecovered)
 	require.False(t, envelope.Data.CacheRefreshed)
 	require.Nil(t, envelope.Data.Quota)
@@ -285,7 +291,7 @@ func TestOpenAIResetQuota_CacheFailureStillRecoversAndReturnsAccount(t *testing.
 	status, envelope := performOpenAIQuotaResetRequest(t, handler)
 
 	require.Equal(t, http.StatusOK, status)
-	require.Equal(t, openAIQuotaResetWarningCacheRefreshFailed, envelope.Data.WarningCode)
+	require.Equal(t, service.OpenAIQuotaResetWarningCacheRefreshFailed, envelope.Data.WarningCode)
 	require.True(t, envelope.Data.AccountStateRecovered)
 	require.False(t, envelope.Data.CacheRefreshed)
 	require.Nil(t, envelope.Data.Quota)
@@ -307,7 +313,7 @@ func TestOpenAIResetQuota_AccountRefreshFailureReportsRecoveredState(t *testing.
 	status, envelope := performOpenAIQuotaResetRequest(t, handler)
 
 	require.Equal(t, http.StatusOK, status)
-	require.Equal(t, openAIQuotaResetWarningAccountRefreshFailed, envelope.Data.WarningCode)
+	require.Equal(t, service.OpenAIQuotaResetWarningAccountRefreshFailed, envelope.Data.WarningCode)
 	require.True(t, envelope.Data.CacheRefreshed)
 	require.True(t, envelope.Data.AccountStateRecovered)
 	require.NotNil(t, envelope.Data.Quota)
@@ -331,7 +337,7 @@ func TestOpenAIResetQuota_CacheAndAccountFailureKeepsFirstWarning(t *testing.T)
 	status, envelope := performOpenAIQuotaResetRequest(t, handler)
 
 	require.Equal(t, http.StatusOK, status)
-	require.Equal(t, openAIQuotaResetWarningCacheRefreshFailed, envelope.Data.WarningCode)
+	require.Equal(t, service.OpenAIQuotaResetWarningCacheRefreshFailed, envelope.Data.WarningCode)
 	require.True(t, envelope.Data.AccountStateRecovered)
 	require.Nil(t, envelope.Data.Account)
 }
@@ -435,7 +441,7 @@ func TestOpenAIQuotaEmptyUsageIsHandledWithoutPanic(t *testing.T) {
 		status, envelope := performOpenAIQuotaResetRequest(t, handler)
 
 		require.Equal(t, http.StatusOK, status)
-		require.Equal(t, openAIQuotaResetWarningCacheRefreshFailed, envelope.Data.WarningCode)
+		require.Equal(t, service.OpenAIQuotaResetWarningCacheRefreshFailed, envelope.Data.WarningCode)
 		require.True(t, envelope.Data.AccountStateRecovered)
 		require.NotNil(t, envelope.Data.Account)
 		require.Zero(t, quota.cacheCalls)

backend/internal/handler/admin/plugin_handler.go

@@ -0,0 +1,257 @@
+package admin
+
+import (
+	"crypto/rand"
+	"encoding/hex"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"io"
+	"mime"
+	"net/http"
+	"os"
+	"path/filepath"
+	"strconv"
+	"strings"
+	"time"
+
+	"github.com/Wei-Shaw/sub2api/internal/pkg/response"
+	"github.com/Wei-Shaw/sub2api/internal/server/middleware"
+	"github.com/Wei-Shaw/sub2api/internal/service"
+	"github.com/gin-gonic/gin"
+)
+
+const pluginUISessionTTL = 30 * time.Minute
+
+// PluginHandler 提供插件安装、生命周期、配置和隔离 UI 资源接口。
+type PluginHandler struct {
+	manager *service.PluginManager
+}
+
+func NewPluginHandler(manager *service.PluginManager) *PluginHandler {
+	return &PluginHandler{manager: manager}
+}
+
+func (h *PluginHandler) List(c *gin.Context) {
+	plugins, err := h.manager.List(c.Request.Context())
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	response.Success(c, plugins)
+}
+
+func (h *PluginHandler) Get(c *gin.Context) {
+	id, ok := pluginIDParam(c)
+	if !ok {
+		return
+	}
+	plugin, err := h.manager.Get(c.Request.Context(), id)
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	response.Success(c, plugin)
+}
+
+func (h *PluginHandler) Upload(c *gin.Context) {
+	maxBytes := h.manager.MaxUploadBytes()
+	c.Request.Body = http.MaxBytesReader(c.Writer, c.Request.Body, maxBytes+(1<<20))
+	file, header, err := c.Request.FormFile("plugin")
+	if err != nil {
+		response.BadRequest(c, "请选择有效的 .s2plugin 文件")
+		return
+	}
+	defer func() { _ = file.Close() }()
+	if !strings.HasSuffix(strings.ToLower(header.Filename), ".s2plugin") {
+		response.BadRequest(c, "插件包扩展名必须是 .s2plugin")
+		return
+	}
+	var installedBy *int64
+	if subject, ok := middleware.GetAuthSubjectFromContext(c); ok && subject.UserID > 0 {
+		userID := subject.UserID
+		installedBy = &userID
+	}
+	plugin, err := h.manager.Install(c.Request.Context(), file, installedBy)
+	if err != nil {
+		response.BadRequest(c, err.Error())
+		return
+	}
+	response.Created(c, plugin)
+}
+
+type pluginEnableRequest struct {
+	AcceptUntested bool `json:"accept_untested"`
+	RolloutPercent int  `json:"rollout_percent"`
+}
+
+func (h *PluginHandler) Enable(c *gin.Context) {
+	id, ok := pluginIDParam(c)
+	if !ok {
+		return
+	}
+	request := pluginEnableRequest{RolloutPercent: 100}
+	if err := c.ShouldBindJSON(&request); err != nil {
+		response.BadRequest(c, "启用参数无效")
+		return
+	}
+	plugin, err := h.manager.Enable(c.Request.Context(), id, request.AcceptUntested, request.RolloutPercent)
+	if err != nil {
+		response.BadRequest(c, err.Error())
+		return
+	}
+	response.Success(c, plugin)
+}
+
+func (h *PluginHandler) Disable(c *gin.Context) {
+	id, ok := pluginIDParam(c)
+	if !ok {
+		return
+	}
+	plugin, err := h.manager.Disable(c.Request.Context(), id)
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	response.Success(c, plugin)
+}
+
+func (h *PluginHandler) Delete(c *gin.Context) {
+	id, ok := pluginIDParam(c)
+	if !ok {
+		return
+	}
+	if err := h.manager.Delete(c.Request.Context(), id); err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	response.Success(c, gin.H{"message": "插件已卸载"})
+}
+
+func (h *PluginHandler) GetConfig(c *gin.Context) {
+	id, ok := pluginIDParam(c)
+	if !ok {
+		return
+	}
+	configJSON, err := h.manager.GetConfig(c.Request.Context(), id)
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	c.Data(http.StatusOK, "application/json; charset=utf-8", configJSON)
+}
+
+func (h *PluginHandler) SaveConfig(c *gin.Context) {
+	id, ok := pluginIDParam(c)
+	if !ok {
+		return
+	}
+	decoder := json.NewDecoder(http.MaxBytesReader(c.Writer, c.Request.Body, 4*1024*1024))
+	decoder.UseNumber()
+	var value any
+	if err := decoder.Decode(&value); err != nil {
+		response.BadRequest(c, "插件配置必须是有效 JSON")
+		return
+	}
+	if err := decoder.Decode(&struct{}{}); err != io.EOF {
+		response.BadRequest(c, "插件配置只能包含一个 JSON 值")
+		return
+	}
+	raw, err := json.Marshal(value)
+	if err != nil {
+		response.BadRequest(c, "插件配置无法序列化")
+		return
+	}
+	saved, err := h.manager.SaveConfig(c.Request.Context(), id, raw)
+	if err != nil {
+		response.BadRequest(c, err.Error())
+		return
+	}
+	c.Data(http.StatusOK, "application/json; charset=utf-8", saved)
+}
+
+func (h *PluginHandler) Test(c *gin.Context) {
+	id, ok := pluginIDParam(c)
+	if !ok {
+		return
+	}
+	result, err := h.manager.Test(c.Request.Context(), id)
+	if err != nil {
+		response.BadRequest(c, err.Error())
+		return
+	}
+	response.Success(c, result)
+}
+
+func (h *PluginHandler) CreateUISession(c *gin.Context) {
+	id, ok := pluginIDParam(c)
+	if !ok {
+		return
+	}
+	assetToken, expires, err := h.manager.CreateUIAssetToken(c.Request.Context(), id, pluginUISessionTTL)
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	bridgeToken, err := randomPluginToken()
+	if err != nil {
+		response.InternalError(c, "创建插件 UI Bridge 失败")
+		return
+	}
+	response.Success(c, gin.H{
+		"url":               fmt.Sprintf("/api/v1/plugin-ui/%s/index.html#bridge_token=%s", assetToken, bridgeToken),
+		"bridge_token":      bridgeToken,
+		"ui_bridge_version": 1,
+		"expires_at":        expires,
+	})
+}
+
+// ServeUIAsset 使用短时随机能力 URL 提供插件静态资源，不向 iframe 暴露管理员凭据。
+func (h *PluginHandler) ServeUIAsset(c *gin.Context) {
+	token := strings.TrimSpace(c.Param("token"))
+	pluginID, err := h.manager.ResolveUIAssetToken(token)
+	if err != nil {
+		c.Status(http.StatusGone)
+		return
+	}
+	relative := strings.TrimPrefix(c.Param("path"), "/")
+	data, logicalPath, err := h.manager.ReadUIAsset(c.Request.Context(), pluginID, relative)
+	if err != nil {
+		if errors.Is(err, os.ErrNotExist) {
+			c.Status(http.StatusNotFound)
+			return
+		}
+		c.Status(http.StatusNotFound)
+		return
+	}
+	contentType := mime.TypeByExtension(filepath.Ext(logicalPath))
+	if contentType == "" {
+		contentType = "application/octet-stream"
+	}
+	c.Header("Cache-Control", "private, no-store")
+	c.Header("Referrer-Policy", "no-referrer")
+	c.Header("X-Content-Type-Options", "nosniff")
+	c.Header("X-Frame-Options", "SAMEORIGIN")
+	// sandbox iframe 没有 allow-same-origin，会以不透明来源加载自己的 CSS/JS。
+	// 资源 URL 由短时随机能力 Token 保护，Bridge Token 只存在于 fragment 中。
+	c.Header("Cross-Origin-Resource-Policy", "cross-origin")
+	c.Header("Content-Security-Policy", "default-src 'none'; script-src 'self' 'unsafe-inline'; style-src 'self' 'unsafe-inline'; img-src 'self' data: blob:; font-src 'self' data:; connect-src 'none'; base-uri 'none'; form-action 'none'; frame-ancestors 'self'; navigate-to 'none'")
+	c.Data(http.StatusOK, contentType, data)
+}
+
+func pluginIDParam(c *gin.Context) (int64, bool) {
+	id, err := strconv.ParseInt(c.Param("id"), 10, 64)
+	if err != nil || id <= 0 {
+		response.BadRequest(c, "插件 ID 无效")
+		return 0, false
+	}
+	return id, true
+}
+
+func randomPluginToken() (string, error) {
+	buffer := make([]byte, 32)
+	if _, err := rand.Read(buffer); err != nil {
+		return "", err
+	}
+	return hex.EncodeToString(buffer), nil
+}

backend/internal/handler/admin/setting_handler.go

@@ -288,6 +288,7 @@ func (h *SettingHandler) GetSettings(c *gin.Context) {
 		MaxClaudeCodeVersion:                                   settings.MaxClaudeCodeVersion,
 		AllowUngroupedKeyScheduling:                            settings.AllowUngroupedKeyScheduling,
 		BackendModeEnabled:                                     settings.BackendModeEnabled,
+		OpenAITTFTMode:                                         settings.OpenAITTFTMode,
 		EnableFingerprintUnification:                           settings.EnableFingerprintUnification,
 		EnableMetadataPassthrough:                              settings.EnableMetadataPassthrough,
 		EnableCCHSigning:                                       settings.EnableCCHSigning,
@@ -382,9 +383,10 @@ func (h *SettingHandler) GetSettings(c *gin.Context) {
 
 		AvailableChannelsEnabled: settings.AvailableChannelsEnabled,
 
-		ModelPlazaEnabled:     settings.ModelPlazaEnabled,
-		ModelPlazaRequireAuth: settings.ModelPlazaRequireAuth,
-		ModelPlazaDescription: settings.ModelPlazaDescription,
+		ModelPlazaEnabled:       settings.ModelPlazaEnabled,
+		ModelPlazaRequireAuth:   settings.ModelPlazaRequireAuth,
+		PluginManagementEnabled: settings.PluginManagementEnabled,
+		ModelPlazaDescription:   settings.ModelPlazaDescription,
 
 		AffiliateEnabled: settings.AffiliateEnabled,
 

backend/internal/handler/admin/setting_handler_audit.go

@@ -452,6 +452,9 @@ func diffSettings(before *service.SystemSettings, after *service.SystemSettings,
 	if before.EnableFingerprintUnification != after.EnableFingerprintUnification {
 		changed = append(changed, "enable_fingerprint_unification")
 	}
+	if before.OpenAITTFTMode != after.OpenAITTFTMode {
+		changed = append(changed, "openai_ttft_mode")
+	}
 	if before.EnableMetadataPassthrough != after.EnableMetadataPassthrough {
 		changed = append(changed, "enable_metadata_passthrough")
 	}

backend/internal/handler/admin/setting_handler_email.go

@@ -17,7 +17,14 @@ type TestSMTPRequest struct {
 	SMTPPort     int    `json:"smtp_port"`
 	SMTPUsername string `json:"smtp_username"`
 	SMTPPassword string `json:"smtp_password"`
-	SMTPUseTLS   bool   `json:"smtp_use_tls"`
+	SMTPUseTLS   *bool  `json:"smtp_use_tls"`
+}
+
+func resolveSMTPUseTLS(requested *bool, savedConfig *service.SMTPConfig) bool {
+	if requested != nil {
+		return *requested
+	}
+	return savedConfig != nil && savedConfig.UseTLS
 }
 
 // TestSMTPConnection 测试SMTP连接
@@ -64,7 +71,7 @@ func (h *SettingHandler) TestSMTPConnection(c *gin.Context) {
 		Port:     req.SMTPPort,
 		Username: req.SMTPUsername,
 		Password: password,
-		UseTLS:   req.SMTPUseTLS,
+		UseTLS:   resolveSMTPUseTLS(req.SMTPUseTLS, savedConfig),
 	}
 
 	err := h.emailService.TestSMTPConnectionWithConfig(config)
@@ -85,7 +92,7 @@ type SendTestEmailRequest struct {
 	SMTPPassword string `json:"smtp_password"`
 	SMTPFrom     string `json:"smtp_from_email"`
 	SMTPFromName string `json:"smtp_from_name"`
-	SMTPUseTLS   bool   `json:"smtp_use_tls"`
+	SMTPUseTLS   *bool  `json:"smtp_use_tls"`
 }
 
 // SendTestEmail 发送测试邮件
@@ -142,7 +149,7 @@ func (h *SettingHandler) SendTestEmail(c *gin.Context) {
 		Password: password,
 		From:     req.SMTPFrom,
 		FromName: req.SMTPFromName,
-		UseTLS:   req.SMTPUseTLS,
+		UseTLS:   resolveSMTPUseTLS(req.SMTPUseTLS, savedConfig),
 	}
 
 	siteName := h.settingService.GetSiteName(c.Request.Context())

backend/internal/handler/admin/setting_handler_email_test.go

@@ -0,0 +1,58 @@
+//go:build unit
+
+package admin
+
+import (
+	"bytes"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+
+	"github.com/Wei-Shaw/sub2api/internal/service"
+
+	"github.com/gin-gonic/gin"
+	"github.com/stretchr/testify/require"
+)
+
+func bindSMTPRequest[T any](t *testing.T, body string) T {
+	t.Helper()
+
+	req := httptest.NewRequest(http.MethodPost, "/", bytes.NewBufferString(body))
+	req.Header.Set("Content-Type", "application/json")
+	c, _ := gin.CreateTestContext(httptest.NewRecorder())
+	c.Request = req
+
+	var payload T
+	require.NoError(t, c.ShouldBindJSON(&payload))
+	return payload
+}
+
+func TestTestSMTPRequestUseTLSOmissionFallsBackToSavedSetting(t *testing.T) {
+	req := bindSMTPRequest[TestSMTPRequest](t, `{}`)
+
+	require.Nil(t, req.SMTPUseTLS)
+	require.True(t, resolveSMTPUseTLS(req.SMTPUseTLS, &service.SMTPConfig{UseTLS: true}))
+}
+
+func TestTestSMTPRequestExplicitFalseOverridesSavedUseTLS(t *testing.T) {
+	req := bindSMTPRequest[TestSMTPRequest](t, `{"smtp_use_tls":false}`)
+
+	require.NotNil(t, req.SMTPUseTLS)
+	require.False(t, resolveSMTPUseTLS(req.SMTPUseTLS, &service.SMTPConfig{UseTLS: true}))
+}
+
+func TestTestSMTPRequestExplicitTrueOverridesMissingSavedUseTLS(t *testing.T) {
+	req := bindSMTPRequest[TestSMTPRequest](t, `{"smtp_use_tls":true}`)
+
+	require.NotNil(t, req.SMTPUseTLS)
+	require.True(t, resolveSMTPUseTLS(req.SMTPUseTLS, nil))
+}
+
+func TestSendTestEmailRequestPreservesUseTLSOmissionSemantics(t *testing.T) {
+	omitted := bindSMTPRequest[SendTestEmailRequest](t, `{"email":"admin@example.com"}`)
+	explicitFalse := bindSMTPRequest[SendTestEmailRequest](t, `{"email":"admin@example.com","smtp_use_tls":false}`)
+	saved := &service.SMTPConfig{UseTLS: true}
+
+	require.True(t, resolveSMTPUseTLS(omitted.SMTPUseTLS, saved))
+	require.False(t, resolveSMTPUseTLS(explicitFalse.SMTPUseTLS, saved))
+}

backend/internal/handler/admin/setting_handler_runtime.go

@@ -154,6 +154,33 @@ func (h *SettingHandler) UpdateRateLimit429CooldownSettings(c *gin.Context) {
 	})
 }
 
+func (h *SettingHandler) GetOpenAIImagesOAuthUnavailableCooldownSettings(c *gin.Context) {
+	settings, err := h.settingService.GetOpenAIImagesOAuthUnavailableCooldownSettings(c.Request.Context())
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	response.Success(c, dto.OpenAIImagesOAuthUnavailableCooldownSettings{CooldownMinutes: settings.CooldownMinutes})
+}
+
+type UpdateOpenAIImagesOAuthUnavailableCooldownSettingsRequest struct {
+	CooldownMinutes int `json:"cooldown_minutes"`
+}
+
+func (h *SettingHandler) UpdateOpenAIImagesOAuthUnavailableCooldownSettings(c *gin.Context) {
+	var req UpdateOpenAIImagesOAuthUnavailableCooldownSettingsRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		response.BadRequest(c, "Invalid request: "+err.Error())
+		return
+	}
+	settings := &service.OpenAIImagesOAuthUnavailableCooldownSettings{CooldownMinutes: req.CooldownMinutes}
+	if err := h.settingService.SetOpenAIImagesOAuthUnavailableCooldownSettings(c.Request.Context(), settings); err != nil {
+		response.BadRequest(c, err.Error())
+		return
+	}
+	response.Success(c, dto.OpenAIImagesOAuthUnavailableCooldownSettings{CooldownMinutes: settings.CooldownMinutes})
+}
+
 // GetPanelRateLimitSettings 获取面板 API 限流配置
 // GET /api/v1/admin/settings/panel-rate-limit
 func (h *SettingHandler) GetPanelRateLimitSettings(c *gin.Context) {

backend/internal/handler/admin/setting_handler_update.go

@@ -243,6 +243,7 @@ type UpdateSettingsRequest struct {
 	BackendModeEnabled bool `json:"backend_mode_enabled"`
 
 	// Gateway forwarding behavior
+	OpenAITTFTMode                         *string `json:"openai_ttft_mode"`
 	EnableFingerprintUnification           *bool   `json:"enable_fingerprint_unification"`
 	EnableMetadataPassthrough              *bool   `json:"enable_metadata_passthrough"`
 	EnableCCHSigning                       *bool   `json:"enable_cch_signing"`
@@ -347,6 +348,9 @@ type UpdateSettingsRequest struct {
 	ModelPlazaRequireAuth *bool   `json:"model_plaza_require_auth"`
 	ModelPlazaDescription *string `json:"model_plaza_description"`
 
+	// Plugin management menu visibility switch; plugin runtime is unaffected.
+	PluginManagementEnabled *bool `json:"plugin_management_enabled"`
+
 	// Affiliate (邀请返利) feature switch
 	AffiliateEnabled *bool `json:"affiliate_enabled"`
 
@@ -438,7 +442,7 @@ func buildSettingKeyByJSONName() map[string]string {
 	out := make(map[string]string, t.NumField())
 	for i := 0; i < t.NumField(); i++ {
 		field := t.Field(i)
-		if field.Type.Kind() == reflect.Ptr {
+		if field.Type.Kind() == reflect.Pointer {
 			continue
 		}
 		name, _, _ := strings.Cut(field.Tag.Get("json"), ",")
@@ -1680,6 +1684,12 @@ func (h *SettingHandler) UpdateSettings(c *gin.Context) {
 			}
 			return previousSettings.EnableFingerprintUnification
 		}(),
+		OpenAITTFTMode: func() string {
+			if req.OpenAITTFTMode != nil {
+				return *req.OpenAITTFTMode
+			}
+			return previousSettings.OpenAITTFTMode
+		}(),
 		EnableMetadataPassthrough: func() bool {
 			if req.EnableMetadataPassthrough != nil {
 				return *req.EnableMetadataPassthrough
@@ -1938,6 +1948,12 @@ func (h *SettingHandler) UpdateSettings(c *gin.Context) {
 			}
 			return previousSettings.ModelPlazaDescription
 		}(),
+		PluginManagementEnabled: func() bool {
+			if req.PluginManagementEnabled != nil {
+				return *req.PluginManagementEnabled
+			}
+			return previousSettings.PluginManagementEnabled
+		}(),
 		AffiliateEnabled: func() bool {
 			if req.AffiliateEnabled != nil {
 				return *req.AffiliateEnabled
@@ -2357,9 +2373,10 @@ func (h *SettingHandler) UpdateSettings(c *gin.Context) {
 
 		AvailableChannelsEnabled: updatedSettings.AvailableChannelsEnabled,
 
-		ModelPlazaEnabled:     updatedSettings.ModelPlazaEnabled,
-		ModelPlazaRequireAuth: updatedSettings.ModelPlazaRequireAuth,
-		ModelPlazaDescription: updatedSettings.ModelPlazaDescription,
+		ModelPlazaEnabled:       updatedSettings.ModelPlazaEnabled,
+		ModelPlazaRequireAuth:   updatedSettings.ModelPlazaRequireAuth,
+		ModelPlazaDescription:   updatedSettings.ModelPlazaDescription,
+		PluginManagementEnabled: updatedSettings.PluginManagementEnabled,
 
 		AffiliateEnabled: updatedSettings.AffiliateEnabled,
 

backend/internal/handler/admin/usage_handler.go

@@ -132,6 +132,12 @@ func (h *UsageHandler) List(c *gin.Context) {
 		stream = &val
 	}
 
+	nativeCompactionV2, err := parseOptionalBoolDashboardFilter(c, "native_compaction_v2")
+	if err != nil {
+		response.BadRequest(c, "Invalid native_compaction_v2 value, use true or false")
+		return
+	}
+
 	var billingType *int8
 	if billingTypeStr := c.Query("billing_type"); billingTypeStr != "" {
 		val, err := strconv.ParseInt(billingTypeStr, 10, 8)
@@ -192,6 +198,7 @@ func (h *UsageHandler) List(c *gin.Context) {
 		ModelFilterSource:     usagestats.ModelSourceRequested,
 		RequestType:           requestType,
 		Stream:                stream,
+		NativeCompactionV2:    nativeCompactionV2,
 		BillingType:           billingType,
 		BillingMode:           billingMode,
 		UpstreamModelMismatch: upstreamModelMismatch,
@@ -276,6 +283,12 @@ func (h *UsageHandler) Stats(c *gin.Context) {
 		stream = &val
 	}
 
+	nativeCompactionV2, err := parseOptionalBoolDashboardFilter(c, "native_compaction_v2")
+	if err != nil {
+		response.BadRequest(c, "Invalid native_compaction_v2 value, use true or false")
+		return
+	}
+
 	var billingType *int8
 	if billingTypeStr := c.Query("billing_type"); billingTypeStr != "" {
 		val, err := strconv.ParseInt(billingTypeStr, 10, 8)
@@ -344,6 +357,7 @@ func (h *UsageHandler) Stats(c *gin.Context) {
 		ModelFilterSource:     usagestats.ModelSourceRequested,
 		RequestType:           requestType,
 		Stream:                stream,
+		NativeCompactionV2:    nativeCompactionV2,
 		BillingType:           billingType,
 		BillingMode:           billingMode,
 		UpstreamModelMismatch: upstreamModelMismatch,

backend/internal/handler/admin/usage_handler_request_type_test.go

@@ -95,6 +95,32 @@ func TestAdminUsageListInvalidStream(t *testing.T) {
 	require.Equal(t, http.StatusBadRequest, rec.Code)
 }
 
+func TestAdminUsageListNativeCompactionFilter(t *testing.T) {
+	repo := &adminUsageRepoCapture{}
+	router := newAdminUsageRequestTypeTestRouter(repo)
+
+	req := httptest.NewRequest(http.MethodGet, "/admin/usage?request_type=stream&native_compaction_v2=true", nil)
+	rec := httptest.NewRecorder()
+	router.ServeHTTP(rec, req)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	require.NotNil(t, repo.listFilters.RequestType)
+	require.Equal(t, int16(service.RequestTypeStream), *repo.listFilters.RequestType)
+	require.NotNil(t, repo.listFilters.NativeCompactionV2)
+	require.True(t, *repo.listFilters.NativeCompactionV2)
+}
+
+func TestAdminUsageListInvalidNativeCompactionFilter(t *testing.T) {
+	repo := &adminUsageRepoCapture{}
+	router := newAdminUsageRequestTypeTestRouter(repo)
+
+	req := httptest.NewRequest(http.MethodGet, "/admin/usage?native_compaction_v2=oops", nil)
+	rec := httptest.NewRecorder()
+	router.ServeHTTP(rec, req)
+
+	require.Equal(t, http.StatusBadRequest, rec.Code)
+}
+
 func TestAdminUsageListExactTotalTrue(t *testing.T) {
 	repo := &adminUsageRepoCapture{}
 	router := newAdminUsageRequestTypeTestRouter(repo)
@@ -144,6 +170,19 @@ func TestAdminUsageStatsRequestTypePriority(t *testing.T) {
 	require.Nil(t, repo.statsFilters.Stream)
 }
 
+func TestAdminUsageStatsNativeCompactionFilter(t *testing.T) {
+	repo := &adminUsageRepoCapture{}
+	router := newAdminUsageRequestTypeTestRouter(repo)
+
+	req := httptest.NewRequest(http.MethodGet, "/admin/usage/stats?native_compaction_v2=true", nil)
+	rec := httptest.NewRecorder()
+	router.ServeHTTP(rec, req)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	require.NotNil(t, repo.statsFilters.NativeCompactionV2)
+	require.True(t, *repo.statsFilters.NativeCompactionV2)
+}
+
 func TestAdminUsageStatsUsesRequestedModelForDisplayModelFilter(t *testing.T) {
 	repo := &adminUsageRepoCapture{}
 	router := newAdminUsageRequestTypeTestRouter(repo)

backend/internal/handler/admin/usage_query_cache.go

@@ -21,6 +21,7 @@ type usageStatsCacheKeyData struct {
 	BillingMode           string `json:"billing_mode"`
 	RequestType           *int16 `json:"request_type"`
 	Stream                *bool  `json:"stream"`
+	NativeCompactionV2    *bool  `json:"native_compaction_v2"`
 	BillingType           *int8  `json:"billing_type"`
 	UpstreamModelMismatch *bool  `json:"upstream_model_mismatch"`
 }
@@ -45,6 +46,7 @@ func usageStatsCacheKey(filters usagestats.UsageLogFilters) string {
 		BillingMode:           filters.BillingMode,
 		RequestType:           filters.RequestType,
 		Stream:                filters.Stream,
+		NativeCompactionV2:    filters.NativeCompactionV2,
 		BillingType:           filters.BillingType,
 		UpstreamModelMismatch: filters.UpstreamModelMismatch,
 	})

backend/internal/handler/admin/user_handler.go

@@ -59,30 +59,32 @@ func NewUserHandler(
 
 // CreateUserRequest represents admin create user request
 type CreateUserRequest struct {
-	Email         string   `json:"email" binding:"required,email"`
-	Password      string   `json:"password" binding:"required,min=6"`
-	Username      string   `json:"username"`
-	Notes         string   `json:"notes"`
-	Role          string   `json:"role" binding:"omitempty,oneof=admin user"`
-	Balance       *float64 `json:"balance"`
-	Concurrency   int      `json:"concurrency"`
-	RPMLimit      int      `json:"rpm_limit"`
-	AllowedGroups []int64  `json:"allowed_groups"`
+	Email                string   `json:"email" binding:"required,email"`
+	Password             string   `json:"password" binding:"required,min=6"`
+	Username             string   `json:"username"`
+	Notes                string   `json:"notes"`
+	Role                 string   `json:"role" binding:"omitempty,oneof=admin user"`
+	Balance              *float64 `json:"balance"`
+	Concurrency          int      `json:"concurrency"`
+	RPMLimit             int      `json:"rpm_limit"`
+	AllowedGroups        []int64  `json:"allowed_groups"`
+	RestrictPublicGroups bool     `json:"restrict_public_groups"`
 }
 
 // UpdateUserRequest represents admin update user request
 // 使用指针类型来区分"未提供"和"设置为0"
 type UpdateUserRequest struct {
-	Email         string   `json:"email" binding:"omitempty,email"`
-	Password      string   `json:"password" binding:"omitempty,min=6"`
-	Username      *string  `json:"username"`
-	Notes         *string  `json:"notes"`
-	Role          string   `json:"role" binding:"omitempty,oneof=admin user"`
-	Balance       *float64 `json:"balance"`
-	Concurrency   *int     `json:"concurrency"`
-	RPMLimit      *int     `json:"rpm_limit"`
-	Status        string   `json:"status" binding:"omitempty,oneof=active disabled"`
-	AllowedGroups *[]int64 `json:"allowed_groups"`
+	Email                string   `json:"email" binding:"omitempty,email"`
+	Password             string   `json:"password" binding:"omitempty,min=6"`
+	Username             *string  `json:"username"`
+	Notes                *string  `json:"notes"`
+	Role                 string   `json:"role" binding:"omitempty,oneof=admin user"`
+	Balance              *float64 `json:"balance"`
+	Concurrency          *int     `json:"concurrency"`
+	RPMLimit             *int     `json:"rpm_limit"`
+	Status               string   `json:"status" binding:"omitempty,oneof=active disabled"`
+	AllowedGroups        *[]int64 `json:"allowed_groups"`
+	RestrictPublicGroups *bool    `json:"restrict_public_groups"`
 	// GroupRates 用户专属分组倍率配置
 	// map[groupID]*rate，nil 表示删除该分组的专属倍率
 	GroupRates map[int64]*float64 `json:"group_rates"`
@@ -284,16 +286,17 @@ func (h *UserHandler) Create(c *gin.Context) {
 	}
 
 	user, err := h.adminService.CreateUser(c.Request.Context(), &service.CreateUserInput{
-		Email:         req.Email,
-		Password:      req.Password,
-		Username:      req.Username,
-		Notes:         req.Notes,
-		Role:          req.Role,
-		Balance:       req.Balance,
-		Concurrency:   req.Concurrency,
-		RPMLimit:      req.RPMLimit,
-		AllowedGroups: req.AllowedGroups,
-		ActorAdminID:  getAdminIDFromContext(c),
+		Email:                req.Email,
+		Password:             req.Password,
+		Username:             req.Username,
+		Notes:                req.Notes,
+		Role:                 req.Role,
+		Balance:              req.Balance,
+		Concurrency:          req.Concurrency,
+		RPMLimit:             req.RPMLimit,
+		AllowedGroups:        req.AllowedGroups,
+		RestrictPublicGroups: req.RestrictPublicGroups,
+		ActorAdminID:         getAdminIDFromContext(c),
 	})
 	if err != nil {
 		response.ErrorFrom(c, err)
@@ -342,18 +345,19 @@ func (h *UserHandler) Update(c *gin.Context) {
 
 	// 使用指针类型直接传递，nil 表示未提供该字段
 	user, err := h.adminService.UpdateUser(c.Request.Context(), userID, &service.UpdateUserInput{
-		Email:         req.Email,
-		Password:      req.Password,
-		Username:      req.Username,
-		Notes:         req.Notes,
-		Role:          req.Role,
-		Balance:       req.Balance,
-		Concurrency:   req.Concurrency,
-		RPMLimit:      req.RPMLimit,
-		Status:        req.Status,
-		AllowedGroups: req.AllowedGroups,
-		GroupRates:    req.GroupRates,
-		ActorAdminID:  getAdminIDFromContext(c),
+		Email:                req.Email,
+		Password:             req.Password,
+		Username:             req.Username,
+		Notes:                req.Notes,
+		Role:                 req.Role,
+		Balance:              req.Balance,
+		Concurrency:          req.Concurrency,
+		RPMLimit:             req.RPMLimit,
+		Status:               req.Status,
+		AllowedGroups:        req.AllowedGroups,
+		RestrictPublicGroups: req.RestrictPublicGroups,
+		GroupRates:           req.GroupRates,
+		ActorAdminID:         getAdminIDFromContext(c),
 	})
 	if err != nil {
 		response.ErrorFrom(c, err)

backend/internal/handler/auth_oidc_oauth.go

@@ -1144,10 +1144,10 @@ func (k oidcJWK) publicKey() (any, error) {
 		if err != nil {
 			return nil, fmt.Errorf("decode ec y: %w", err)
 		}
-		if !curve.IsOnCurve(x, y) {
+		if !curve.IsOnCurve(x, y) { //nolint:staticcheck // JWK 以裸坐标给出公钥；替换为 ecdsa.ParseUncompressedPublicKey 需改变点编码，待单独迁移
 			return nil, errors.New("ec point is not on curve")
 		}
-		return &ecdsa.PublicKey{Curve: curve, X: x, Y: y}, nil
+		return &ecdsa.PublicKey{Curve: curve, X: x, Y: y}, nil //nolint:staticcheck // 同上
 	default:
 		return nil, fmt.Errorf("unsupported jwk kty: %s", k.Kty)
 	}

backend/internal/handler/available_channel_handler.go

@@ -284,13 +284,13 @@ func toUserSupportedModels(
 	return out
 }
 
-// toUserPricing 将 service 层定价转换为用户 DTO；入参为 nil 时返回 nil。
-func toUserPricing(p *service.ChannelModelPricing) *userSupportedModelPricing {
-	if p == nil {
+// toUserPricingIntervals 将定价区间转换为用户 DTO 白名单形态；nil 入参返回 nil（JSON omitempty 可省略）。
+func toUserPricingIntervals(src []service.PricingInterval) []userPricingIntervalDTO {
+	if src == nil {
 		return nil
 	}
-	intervals := make([]userPricingIntervalDTO, 0, len(p.Intervals))
-	for _, iv := range p.Intervals {
+	intervals := make([]userPricingIntervalDTO, 0, len(src))
+	for _, iv := range src {
 		intervals = append(intervals, userPricingIntervalDTO{
 			MinTokens:       iv.MinTokens,
 			MaxTokens:       iv.MaxTokens,
@@ -302,6 +302,19 @@ func toUserPricing(p *service.ChannelModelPricing) *userSupportedModelPricing {
 			PerRequestPrice: iv.PerRequestPrice,
 		})
 	}
+	return intervals
+}
+
+// toUserPricing 将 service 层定价转换为用户 DTO；入参为 nil 时返回 nil。
+func toUserPricing(p *service.ChannelModelPricing) *userSupportedModelPricing {
+	if p == nil {
+		return nil
+	}
+	intervals := toUserPricingIntervals(p.Intervals)
+	if intervals == nil {
+		// 用户侧定价的 intervals 固定输出数组（空配置为 []），保持既有契约。
+		intervals = []userPricingIntervalDTO{}
+	}
 	billingMode := string(p.BillingMode)
 	if billingMode == "" {
 		billingMode = string(service.BillingModeToken)

backend/internal/handler/channel_monitor_v2_handler.go

@@ -1,6 +1,7 @@
 package handler
 
 import (
+	"context"
 	"errors"
 	"net/http"
 	"strconv"
@@ -13,11 +14,16 @@ import (
 )
 
 type ChannelMonitorV2Handler struct {
-	service *service.ChannelMonitorV2Service
+	service       *service.ChannelMonitorV2Service
+	apiKeyService channelMonitorV2GroupAuthorizer
 }
 
-func NewChannelMonitorV2Handler(svc *service.ChannelMonitorV2Service) *ChannelMonitorV2Handler {
-	return &ChannelMonitorV2Handler{service: svc}
+type channelMonitorV2GroupAuthorizer interface {
+	GetAvailableGroups(ctx context.Context, userID int64) ([]service.Group, error)
+}
+
+func NewChannelMonitorV2Handler(svc *service.ChannelMonitorV2Service, apiKeyService *service.APIKeyService) *ChannelMonitorV2Handler {
+	return &ChannelMonitorV2Handler{service: svc, apiKeyService: apiKeyService}
 }
 
 // channelMonitorV2IsAdmin is true when the request already passed admin auth
@@ -68,13 +74,17 @@ func (h *ChannelMonitorV2Handler) Dimensions(c *gin.Context) {
 	if !ok {
 		return
 	}
+	admin := channelMonitorV2IsAdmin(c)
+	if !h.scopeFilter(c, &filter, admin) {
+		return
+	}
 	result, err := h.service.Dimensions(c.Request.Context(), filter)
 	if err != nil {
 		response.ErrorFrom(c, err)
 		return
 	}
 	// Admin and user share this handler; only non-admin responses strip volume.
-	if !channelMonitorV2IsAdmin(c) {
+	if !admin {
 		service.RedactChannelMonitorV2Dimensions(result)
 	}
 	response.Success(c, result)
@@ -94,6 +104,9 @@ func (h *ChannelMonitorV2Handler) snapshot(c *gin.Context, admin bool) {
 	if !ok {
 		return
 	}
+	if !h.scopeFilter(c, &filter, admin) {
+		return
+	}
 	result, err := h.service.Snapshot(c.Request.Context(), filter, admin)
 	if err != nil {
 		response.ErrorFrom(c, err)
@@ -107,6 +120,9 @@ func (h *ChannelMonitorV2Handler) models(c *gin.Context, admin bool) {
 	if !ok {
 		return
 	}
+	if !h.scopeFilter(c, &filter, admin) {
+		return
+	}
 	result, err := h.service.Models(c.Request.Context(), filter, admin)
 	if err != nil {
 		response.ErrorFrom(c, err)
@@ -125,6 +141,9 @@ func (h *ChannelMonitorV2Handler) matrix(c *gin.Context, admin bool) {
 		response.BadRequest(c, err.Error())
 		return
 	}
+	if !h.scopeFilter(c, &filter, admin) {
+		return
+	}
 	result, err := h.service.Matrix(c.Request.Context(), filter, groupBy, admin)
 	if err != nil {
 		response.ErrorFrom(c, err)
@@ -138,7 +157,11 @@ func (h *ChannelMonitorV2Handler) Errors(c *gin.Context) {
 	if !ok {
 		return
 	}
-	result, err := h.service.ErrorsForViewer(c.Request.Context(), filter, channelMonitorV2IsAdmin(c))
+	admin := channelMonitorV2IsAdmin(c)
+	if !h.scopeFilter(c, &filter, admin) {
+		return
+	}
+	result, err := h.service.ErrorsForViewer(c.Request.Context(), filter, admin)
 	if err != nil {
 		response.ErrorFrom(c, err)
 		return
@@ -156,6 +179,9 @@ func (h *ChannelMonitorV2Handler) users(c *gin.Context, admin bool) {
 		response.Error(c, http.StatusUnauthorized, "user not found in context")
 		return
 	}
+	if !h.scopeFilter(c, &filter, admin) {
+		return
+	}
 	result, err := h.service.Users(c.Request.Context(), filter, subject.UserID, admin)
 	if err != nil {
 		response.ErrorFrom(c, err)
@@ -164,6 +190,32 @@ func (h *ChannelMonitorV2Handler) users(c *gin.Context, admin bool) {
 	response.Success(c, result)
 }
 
+func (h *ChannelMonitorV2Handler) scopeFilter(c *gin.Context, filter *service.ChannelMonitorV2Filter, admin bool) bool {
+	if admin {
+		return true
+	}
+	if h.apiKeyService == nil {
+		response.Error(c, http.StatusInternalServerError, "channel monitor group authorization unavailable")
+		return false
+	}
+	subject, ok := middleware.GetAuthSubjectFromContext(c)
+	if !ok || subject.UserID <= 0 {
+		response.Unauthorized(c, "user not found in context")
+		return false
+	}
+	groups, err := h.apiKeyService.GetAvailableGroups(c.Request.Context(), subject.UserID)
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return false
+	}
+	filter.RestrictGroups = true
+	filter.AllowedGroupIDs = make([]int64, 0, len(groups))
+	for i := range groups {
+		filter.AllowedGroupIDs = append(filter.AllowedGroupIDs, groups[i].ID)
+	}
+	return true
+}
+
 func (h *ChannelMonitorV2Handler) parseFilter(c *gin.Context) (service.ChannelMonitorV2Filter, bool) {
 	groups, err := parseChannelMonitorV2GroupIDs(queryList(c, "group_id"))
 	if err != nil {

backend/internal/handler/channel_monitor_v2_handler_test.go

@@ -1,15 +1,28 @@
 package handler
 
 import (
+	"context"
 	"net/http"
 	"net/http/httptest"
 	"testing"
 
+	"github.com/Wei-Shaw/sub2api/internal/server/middleware"
 	"github.com/Wei-Shaw/sub2api/internal/service"
 	"github.com/gin-gonic/gin"
 	"github.com/stretchr/testify/require"
 )
 
+type channelMonitorV2GroupAuthorizerStub struct {
+	groups []service.Group
+	err    error
+	calls  []int64
+}
+
+func (s *channelMonitorV2GroupAuthorizerStub) GetAvailableGroups(_ context.Context, userID int64) ([]service.Group, error) {
+	s.calls = append(s.calls, userID)
+	return s.groups, s.err
+}
+
 func TestChannelMonitorV2QueryListSupportsRepeatedAndCommaValues(t *testing.T) {
 	c, _ := gin.CreateTestContext(nil)
 	c.Request = httptest.NewRequest("GET", "/?platform=openai,grok&platform=anthropic", nil)
@@ -29,7 +42,46 @@ func TestChannelMonitorV2MatrixHandlerRejectsInvalidGroupBy(t *testing.T) {
 	recorder := httptest.NewRecorder()
 	c, _ := gin.CreateTestContext(recorder)
 	c.Request = httptest.NewRequest(http.MethodGet, "/channel-monitor-v2/matrix?group_by=invalid", nil)
-	h := NewChannelMonitorV2Handler(service.NewChannelMonitorV2Service(nil))
+	h := NewChannelMonitorV2Handler(service.NewChannelMonitorV2Service(nil), nil)
 	h.Matrix(c)
 	require.Equal(t, http.StatusBadRequest, recorder.Code)
 }
+
+func TestChannelMonitorV2ScopeFilterUsesAvailableGroupsForOrdinaryUser(t *testing.T) {
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+	c.Request = httptest.NewRequest(http.MethodGet, "/channel-monitor-v2/snapshot", nil)
+	c.Set(string(middleware.ContextKeyUser), middleware.AuthSubject{UserID: 42})
+	authorizer := &channelMonitorV2GroupAuthorizerStub{groups: []service.Group{{ID: 3}, {ID: 7}}}
+	h := &ChannelMonitorV2Handler{apiKeyService: authorizer}
+	filter := service.ChannelMonitorV2Filter{GroupIDs: []int64{7, 9}}
+
+	require.True(t, h.scopeFilter(c, &filter, false))
+	require.True(t, filter.RestrictGroups)
+	require.Equal(t, []int64{3, 7}, filter.AllowedGroupIDs)
+	require.Equal(t, []int64{42}, authorizer.calls)
+}
+
+func TestChannelMonitorV2ScopeFilterPreservesEmptyOrdinaryUserScope(t *testing.T) {
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+	c.Request = httptest.NewRequest(http.MethodGet, "/channel-monitor-v2/snapshot", nil)
+	c.Set(string(middleware.ContextKeyUser), middleware.AuthSubject{UserID: 42})
+	h := &ChannelMonitorV2Handler{apiKeyService: &channelMonitorV2GroupAuthorizerStub{}}
+	filter := service.ChannelMonitorV2Filter{}
+
+	require.True(t, h.scopeFilter(c, &filter, false))
+	require.True(t, filter.RestrictGroups)
+	require.Empty(t, filter.AllowedGroupIDs)
+}
+
+func TestChannelMonitorV2ScopeFilterLeavesAdminUnrestricted(t *testing.T) {
+	authorizer := &channelMonitorV2GroupAuthorizerStub{}
+	h := &ChannelMonitorV2Handler{apiKeyService: authorizer}
+	filter := service.ChannelMonitorV2Filter{GroupIDs: []int64{9}}
+
+	require.True(t, h.scopeFilter(nil, &filter, true))
+	require.False(t, filter.RestrictGroups)
+	require.Nil(t, filter.AllowedGroupIDs)
+	require.Empty(t, authorizer.calls)
+}

backend/internal/handler/composite_platform.go

@@ -72,7 +72,36 @@ func openAIReasoningEffortPolicyForRequest(c *gin.Context, apiKey *service.APIKe
 	return apiKey.Group.MaxReasoningEffort, apiKey.Group.ReasoningEffortMappings, true
 }
 
+func bindRequestedReasoningEffort(c *gin.Context, body []byte, model string) {
+	if c == nil || c.Request == nil {
+		return
+	}
+	effort := service.CanonicalRequestedReasoningEffort(body, model)
+	if effort == nil {
+		return
+	}
+	c.Request = c.Request.WithContext(service.WithRequestedReasoningEffort(c.Request.Context(), *effort))
+}
+
+func stampOpenAIRequestedReasoningEffort(result *service.OpenAIForwardResult, c *gin.Context) {
+	if result == nil || result.RequestedReasoningEffort != nil {
+		return
+	}
+	if c == nil || c.Request == nil {
+		return
+	}
+	result.RequestedReasoningEffort = service.RequestedReasoningEffortFromContext(c.Request.Context())
+}
+
+func stampForwardRequestedReasoningEffort(result *service.ForwardResult, requested *string) {
+	if result == nil || result.RequestedReasoningEffort != nil {
+		return
+	}
+	result.RequestedReasoningEffort = requested
+}
+
 func applyOpenAIReasoningEffortPolicyForRequest(c *gin.Context, apiKey *service.APIKey, body []byte) ([]byte, bool) {
+	bindRequestedReasoningEffort(c, body, strings.TrimSpace(gjson.GetBytes(body, "model").String()))
 	maxEffort, mappings, ok := openAIReasoningEffortPolicyForRequest(c, apiKey)
 	if !ok {
 		return body, false
@@ -84,6 +113,7 @@ func bindOpenAIReasoningEffortPolicyForMessagesRequest(c *gin.Context, apiKey *s
 	if c == nil || c.Request == nil {
 		return
 	}
+	bindRequestedReasoningEffort(c, body, strings.TrimSpace(gjson.GetBytes(body, "model").String()))
 	// The Messages bridge synthesizes a default OpenAI effort when
 	// output_config.effort is omitted. Bind the group policy only for an
 	// explicit client value so the ceiling does not alter that default.

backend/internal/handler/composite_platform_test.go

@@ -22,18 +22,43 @@ func TestCompositeTargetPlatformAllowedResolvesKnownAllowedModel(t *testing.T) {
 	require.Equal(t, service.PlatformOpenAI, platform)
 }
 
-func TestOpenAICompatibleTextTargetAllowsCompositeGrokModel(t *testing.T) {
+func TestOpenAICompatibleTextTargetAllowsCompositeProviders(t *testing.T) {
 	gin.SetMode(gin.TestMode)
 
-	for _, path := range []string{"/v1/messages", "/v1/chat/completions"} {
-		c, _ := gin.CreateTestContext(httptest.NewRecorder())
-		c.Request = httptest.NewRequest("POST", path, nil)
-		apiKey := &service.APIKey{Group: &service.Group{Platform: service.PlatformComposite}}
+	providers := []struct {
+		model    string
+		platform string
+	}{
+		{model: "grok-4.3", platform: service.PlatformGrok},
+		{model: "kimi-k2-thinking", platform: service.PlatformKimi},
+		{model: "k3", platform: service.PlatformKimi},
+		{model: "glm-5.2", platform: service.PlatformZhipu},
+		{model: "deepseek-v3.2", platform: service.PlatformDeepseek},
+	}
+	for _, path := range []string{"/v1/messages", "/v1/chat/completions", "/v1/responses", "/v1/responses/input_tokens", "/v1/messages/count_tokens"} {
+		for _, provider := range providers {
+			c, _ := gin.CreateTestContext(httptest.NewRecorder())
+			c.Request = httptest.NewRequest("POST", path, nil)
+			apiKey := &service.APIKey{Group: &service.Group{Platform: service.PlatformComposite}}
 
-		require.True(t, openAICompatibleTextTargetAllowed(c, apiKey, "grok-4.3"), "path=%s", path)
-		platform, ok := service.ResolvedTargetPlatformFromContext(c.Request.Context())
-		require.True(t, ok, "path=%s", path)
-		require.Equal(t, service.PlatformGrok, platform, "path=%s", path)
+			require.True(t, openAICompatibleTextTargetAllowed(c, apiKey, provider.model), "path=%s model=%s", path, provider.model)
+			platform, ok := service.ResolvedTargetPlatformFromContext(c.Request.Context())
+			require.True(t, ok, "path=%s model=%s", path, provider.model)
+			require.Equal(t, provider.platform, platform, "path=%s model=%s", path, provider.model)
+		}
+	}
+}
+
+// WS ingress 对 CN 账号既过不了 transport 过滤、HTTP 桥也没有 Responses 转换，
+// 放行只会把明确的策略拒绝换成 "no available account"，因此 WS 白名单保持 openai+grok。
+func TestResponsesWebSocketCompositePlatformGuardKeepsOpenAIAndGrokOnly(t *testing.T) {
+	require.True(t, isResponsesWebSocketCompositePlatform(service.PlatformOpenAI))
+	require.True(t, isResponsesWebSocketCompositePlatform(service.PlatformGrok))
+	for _, platform := range []string{
+		service.PlatformKimi, service.PlatformZhipu, service.PlatformDeepseek,
+		service.PlatformAnthropic, service.PlatformGemini,
+	} {
+		require.False(t, isResponsesWebSocketCompositePlatform(platform), "platform=%s", platform)
 	}
 }
 
@@ -95,6 +120,9 @@ func TestOpenAIReasoningEffortPolicyForCompositeTarget(t *testing.T) {
 	got, changed := applyOpenAIReasoningEffortPolicyForRequest(openAICtx, apiKey, body)
 	require.True(t, changed)
 	require.JSONEq(t, `{"reasoning":{"effort":"medium"}}`, string(got))
+	requested := service.RequestedReasoningEffortFromContext(openAICtx.Request.Context())
+	require.NotNil(t, requested)
+	require.Equal(t, "max", *requested)
 
 	bindOpenAIReasoningEffortPolicyForMessagesRequest(openAICtx, apiKey, []byte(`{"output_config":{"effort":"max"}}`))
 	bound, changed := service.ApplyOpenAIReasoningEffortPolicyFromContext(openAICtx.Request.Context(), body)

backend/internal/handler/dto/mappers.go

@@ -3,6 +3,7 @@ package dto
 
 import (
 	"strconv"
+	"strings"
 	"time"
 
 	"github.com/Wei-Shaw/sub2api/internal/service"
@@ -68,10 +69,11 @@ func UserFromServiceAdmin(u *service.User) *AdminUser {
 		return nil
 	}
 	return &AdminUser{
-		User:       *base,
-		Notes:      u.Notes,
-		LastUsedAt: u.LastUsedAt,
-		GroupRates: u.GroupRates,
+		User:                 *base,
+		Notes:                u.Notes,
+		LastUsedAt:           u.LastUsedAt,
+		GroupRates:           u.GroupRates,
+		RestrictPublicGroups: u.RestrictPublicGroups,
 	}
 }
 
@@ -643,7 +645,7 @@ func usageLogFromServiceUser(l *service.UsageLog) UsageLog {
 		RequestID:                 l.RequestID,
 		Model:                     requestedModel,
 		ServiceTier:               l.ServiceTier,
-		ReasoningEffort:           l.ReasoningEffort,
+		ReasoningEffort:           userFacingReasoningEffort(l),
 		InboundEndpoint:           l.InboundEndpoint,
 		GroupID:                   l.GroupID,
 		SubscriptionID:            l.SubscriptionID,
@@ -665,6 +667,7 @@ func usageLogFromServiceUser(l *service.UsageLog) UsageLog {
 		RequestType:               requestType.String(),
 		Stream:                    stream,
 		OpenAIWSMode:              openAIWSMode,
+		NativeCompactionV2:        l.NativeCompactionV2,
 		DurationMs:                l.DurationMs,
 		FirstTokenMs:              l.FirstTokenMs,
 		ImageCount:                l.ImageCount,
@@ -710,20 +713,53 @@ func UsageLogFromServiceAdmin(l *service.UsageLog) *AdminUsageLog {
 	usageLog := usageLogFromServiceUser(l)
 	usageLog.UpstreamEndpoint = l.UpstreamEndpoint
 	return &AdminUsageLog{
-		UsageLog:              usageLog,
-		UpstreamModel:         l.UpstreamModel,
-		UpstreamResponseModel: l.UpstreamResponseModel,
-		UpstreamModelMismatch: l.UpstreamModelMismatch,
-		ChannelID:             l.ChannelID,
-		ModelMappingChain:     l.ModelMappingChain,
-		BillingTier:           l.BillingTier,
-		AccountRateMultiplier: l.AccountRateMultiplier,
-		AccountStatsCost:      l.AccountStatsCost,
-		IPAddress:             l.IPAddress,
-		Account:               AccountSummaryFromService(l.Account),
+		UsageLog:                usageLog,
+		UpstreamModel:           l.UpstreamModel,
+		UpstreamReasoningEffort: adminUpstreamReasoningEffort(l),
+		UpstreamResponseModel:   l.UpstreamResponseModel,
+		UpstreamModelMismatch:   l.UpstreamModelMismatch,
+		ChannelID:               l.ChannelID,
+		ModelMappingChain:       l.ModelMappingChain,
+		BillingTier:             l.BillingTier,
+		AccountRateMultiplier:   l.AccountRateMultiplier,
+		AccountStatsCost:        l.AccountStatsCost,
+		IPAddress:               l.IPAddress,
+		Account:                 AccountSummaryFromService(l.Account),
 	}
 }
 
+func userFacingReasoningEffort(l *service.UsageLog) *string {
+	if l == nil {
+		return nil
+	}
+	if requested := strings.TrimSpace(derefString(l.RequestedReasoningEffort)); requested != "" {
+		return &requested
+	}
+	return l.ReasoningEffort
+}
+
+func adminUpstreamReasoningEffort(l *service.UsageLog) *string {
+	if l == nil {
+		return nil
+	}
+	forwarded := strings.TrimSpace(derefString(l.ReasoningEffort))
+	if forwarded == "" {
+		return nil
+	}
+	requested := userFacingReasoningEffort(l)
+	if requested != nil && service.NormalizeMaxReasoningEffort(*requested) == service.NormalizeMaxReasoningEffort(forwarded) {
+		return nil
+	}
+	return &forwarded
+}
+
+func derefString(value *string) string {
+	if value == nil {
+		return ""
+	}
+	return *value
+}
+
 func UsageCleanupTaskFromService(task *service.UsageCleanupTask) *UsageCleanupTask {
 	if task == nil {
 		return nil

backend/internal/handler/dto/mappers_usage_test.go

@@ -28,6 +28,27 @@ func TestUsageLogFromService_IncludesOpenAIWSMode(t *testing.T) {
 	require.False(t, UsageLogFromServiceAdmin(httpLog).OpenAIWSMode)
 }
 
+func TestUsageLogFromService_PreservesNativeCompactionAndStream(t *testing.T) {
+	t.Parallel()
+
+	log := &service.UsageLog{
+		RequestID:          "resp_compaction",
+		Model:              "gpt-5.6-sol",
+		RequestType:        service.RequestTypeStream,
+		Stream:             true,
+		NativeCompactionV2: true,
+	}
+
+	userDTO := UsageLogFromService(log)
+	adminDTO := UsageLogFromServiceAdmin(log)
+	require.Equal(t, "stream", userDTO.RequestType)
+	require.True(t, userDTO.Stream)
+	require.True(t, userDTO.NativeCompactionV2)
+	require.Equal(t, "stream", adminDTO.RequestType)
+	require.True(t, adminDTO.Stream)
+	require.True(t, adminDTO.NativeCompactionV2)
+}
+
 func TestUsageLogFromService_PrefersRequestTypeForLegacyFields(t *testing.T) {
 	t.Parallel()
 
@@ -179,6 +200,61 @@ func TestUsageLogFromService_KeepsUserBillingAndIPWithoutAdminCostFields(t *test
 	require.NotContains(t, string(userJSON), "account_cost")
 }
 
+func TestUsageLogFromService_UsersSeeRequestedReasoningEffortOnly(t *testing.T) {
+	t.Parallel()
+
+	requested := "max"
+	forwarded := "xhigh"
+	log := &service.UsageLog{
+		RequestID:                "req_effort",
+		Model:                    "gpt-5.4",
+		ReasoningEffort:          &forwarded,
+		RequestedReasoningEffort: &requested,
+	}
+
+	userDTO := UsageLogFromService(log)
+	adminDTO := UsageLogFromServiceAdmin(log)
+
+	require.NotNil(t, userDTO.ReasoningEffort)
+	require.Equal(t, requested, *userDTO.ReasoningEffort)
+	require.NotNil(t, adminDTO.ReasoningEffort)
+	require.Equal(t, requested, *adminDTO.ReasoningEffort)
+	require.NotNil(t, adminDTO.UpstreamReasoningEffort)
+	require.Equal(t, forwarded, *adminDTO.UpstreamReasoningEffort)
+
+	userJSON, err := json.Marshal(userDTO)
+	require.NoError(t, err)
+	require.Contains(t, string(userJSON), `"reasoning_effort":"max"`)
+	require.NotContains(t, string(userJSON), "upstream_reasoning_effort")
+	require.NotContains(t, string(userJSON), "requested_reasoning_effort")
+
+	adminJSON, err := json.Marshal(adminDTO)
+	require.NoError(t, err)
+	require.Contains(t, string(adminJSON), `"reasoning_effort":"max"`)
+	require.Contains(t, string(adminJSON), `"upstream_reasoning_effort":"xhigh"`)
+}
+
+func TestUsageLogFromService_OmitsUpstreamReasoningEffortWhenUnmapped(t *testing.T) {
+	t.Parallel()
+
+	effort := "high"
+	log := &service.UsageLog{
+		RequestID:                "req_effort_same",
+		Model:                    "gpt-5.4",
+		ReasoningEffort:          &effort,
+		RequestedReasoningEffort: &effort,
+	}
+
+	adminDTO := UsageLogFromServiceAdmin(log)
+	require.NotNil(t, adminDTO.ReasoningEffort)
+	require.Equal(t, effort, *adminDTO.ReasoningEffort)
+	require.Nil(t, adminDTO.UpstreamReasoningEffort)
+
+	adminJSON, err := json.Marshal(adminDTO)
+	require.NoError(t, err)
+	require.NotContains(t, string(adminJSON), "upstream_reasoning_effort")
+}
+
 func TestUsageLogFromService_FallsBackToLegacyModelWhenRequestedModelMissing(t *testing.T) {
 	t.Parallel()
 

backend/internal/handler/dto/public_settings_injection_schema_test.go

@@ -29,8 +29,6 @@ func TestPublicSettingsInjectionPayload_SchemaDoesNotDrift(t *testing.T) {
 
 	// Fields that legitimately live only on the DTO. Keep tiny; document each.
 	dtoOnlyFields := map[string]string{
-		// sora_client_enabled is an upstream-only field the fork does not surface.
-		"sora_client_enabled": "upstream-only field, not used on this fork",
 		// force_email_on_third_party_signup lives on the DTO but is not injected via SSR.
 		"force_email_on_third_party_signup": "auth-source default, not a feature flag",
 	}

backend/internal/handler/dto/settings.go

@@ -201,6 +201,7 @@ type SystemSettings struct {
 	BackendModeEnabled bool `json:"backend_mode_enabled"`
 
 	// Gateway forwarding behavior
+	OpenAITTFTMode                         string `json:"openai_ttft_mode"`
 	EnableFingerprintUnification           bool   `json:"enable_fingerprint_unification"`
 	EnableMetadataPassthrough              bool   `json:"enable_metadata_passthrough"`
 	EnableCCHSigning                       bool   `json:"enable_cch_signing"`
@@ -316,9 +317,10 @@ type SystemSettings struct {
 	AvailableChannelsEnabled bool `json:"available_channels_enabled"`
 
 	// Model Plaza feature (public group/model pricing showcase)
-	ModelPlazaEnabled     bool   `json:"model_plaza_enabled"`
-	ModelPlazaRequireAuth bool   `json:"model_plaza_require_auth"`
-	ModelPlazaDescription string `json:"model_plaza_description"`
+	ModelPlazaEnabled       bool   `json:"model_plaza_enabled"`
+	ModelPlazaRequireAuth   bool   `json:"model_plaza_require_auth"`
+	ModelPlazaDescription   string `json:"model_plaza_description"`
+	PluginManagementEnabled bool   `json:"plugin_management_enabled"`
 
 	// 风控中心功能开关
 	RiskControlEnabled bool `json:"risk_control_enabled"`
@@ -398,7 +400,6 @@ type PublicSettings struct {
 	OIDCOAuthProviderName               string                   `json:"oidc_oauth_provider_name"`
 	GitHubOAuthEnabled                  bool                     `json:"github_oauth_enabled"`
 	GoogleOAuthEnabled                  bool                     `json:"google_oauth_enabled"`
-	SoraClientEnabled                   bool                     `json:"sora_client_enabled"`
 	BackendModeEnabled                  bool                     `json:"backend_mode_enabled"`
 	PaymentEnabled                      bool                     `json:"payment_enabled"`
 	Version                             string                   `json:"version"`
@@ -419,8 +420,9 @@ type PublicSettings struct {
 
 	AvailableChannelsEnabled bool `json:"available_channels_enabled"`
 
-	ModelPlazaEnabled     bool `json:"model_plaza_enabled"`
-	ModelPlazaRequireAuth bool `json:"model_plaza_require_auth"`
+	ModelPlazaEnabled       bool `json:"model_plaza_enabled"`
+	ModelPlazaRequireAuth   bool `json:"model_plaza_require_auth"`
+	PluginManagementEnabled bool `json:"plugin_management_enabled"`
 
 	AffiliateEnabled bool `json:"affiliate_enabled"`
 
@@ -447,6 +449,10 @@ type RateLimit429CooldownSettings struct {
 	CooldownSeconds int  `json:"cooldown_seconds"`
 }
 
+type OpenAIImagesOAuthUnavailableCooldownSettings struct {
+	CooldownMinutes int `json:"cooldown_minutes"`
+}
+
 // PanelRateLimitSettings 面板 API 限流配置 DTO
 type PanelRateLimitSettings struct {
 	Enabled     bool `json:"enabled"`

backend/internal/handler/dto/types.go

@@ -48,6 +48,9 @@ type AdminUser struct {
 	// GroupRates 用户专属分组倍率配置
 	// map[groupID]rateMultiplier
 	GroupRates map[int64]float64 `json:"group_rates,omitempty"`
+	// RestrictPublicGroups 为 true 时，该用户仅可使用 allowed_groups 中列出的
+	// 公开分组。这是管理侧的权限开关，不下发给用户自身的接口。
+	RestrictPublicGroups bool `json:"restrict_public_groups"`
 }
 
 type APIKey struct {
@@ -486,8 +489,9 @@ type UsageLog struct {
 	Model     string `json:"model"`
 	// ServiceTier records the OpenAI service tier used for billing, e.g. "priority" / "flex".
 	ServiceTier *string `json:"service_tier,omitempty"`
-	// ReasoningEffort is the request's reasoning effort level.
+	// ReasoningEffort is the client-requested effort (mapping-hidden, like Model).
 	// OpenAI: "low"/"medium"/"high"/"xhigh"; Claude: "low"/"medium"/"high"/"max".
+	// Historical rows without requested_reasoning_effort fall back to the stored effective value.
 	ReasoningEffort *string `json:"reasoning_effort,omitempty"`
 	// InboundEndpoint is the client-facing API endpoint path, e.g. /v1/chat/completions.
 	InboundEndpoint *string `json:"inbound_endpoint,omitempty"`
@@ -518,8 +522,11 @@ type UsageLog struct {
 	RequestType  string `json:"request_type"`
 	Stream       bool   `json:"stream"`
 	OpenAIWSMode bool   `json:"openai_ws_mode"`
-	DurationMs   *int   `json:"duration_ms"`
-	FirstTokenMs *int   `json:"first_token_ms"`
+	// NativeCompactionV2 is true only for requests positively identified at
+	// runtime as the native OpenAI remote compaction v2 wire.
+	NativeCompactionV2 bool `json:"native_compaction_v2"`
+	DurationMs         *int `json:"duration_ms"`
+	FirstTokenMs       *int `json:"first_token_ms"`
 
 	// 图片生成字段
 	ImageCount         int            `json:"image_count"`
@@ -563,6 +570,9 @@ type AdminUsageLog struct {
 	// UpstreamModel is the actual model sent to the upstream provider after mapping.
 	// Omitted when no mapping was applied (requested model was used as-is).
 	UpstreamModel *string `json:"upstream_model,omitempty"`
+	// UpstreamReasoningEffort is the effort actually forwarded after group policy /
+	// model-family remapping. Omitted when it matches the client-requested value.
+	UpstreamReasoningEffort *string `json:"upstream_reasoning_effort,omitempty"`
 	// UpstreamResponseModel is the raw model declared by the upstream response.
 	UpstreamResponseModel *string `json:"upstream_response_model,omitempty"`
 	// UpstreamModelMismatch is nil when the upstream did not declare a model.

backend/internal/handler/endpoint.go

@@ -15,20 +15,21 @@ import (
 // ──────────────────────────────────────────────────────────
 
 const (
-	EndpointMessages          = "/v1/messages"
-	EndpointChatCompletions   = "/v1/chat/completions"
-	EndpointEmbeddings        = "/v1/embeddings"
-	EndpointAlphaSearch       = "/v1/alpha/search"
-	EndpointResponses         = "/v1/responses"
-	EndpointResponsesCompact  = "/v1/responses/compact"
-	EndpointImagesGenerations = "/v1/images/generations"
-	EndpointImagesEdits       = "/v1/images/edits"
-	EndpointImageTasks        = "/v1/images/tasks"
-	EndpointVideosGenerations = "/v1/videos/generations"
-	EndpointVideosEdits       = "/v1/videos/edits"
-	EndpointVideosExtensions  = "/v1/videos/extensions"
-	EndpointVideos            = "/v1/videos"
-	EndpointGeminiModels      = "/v1beta/models"
+	EndpointMessages             = "/v1/messages"
+	EndpointChatCompletions      = "/v1/chat/completions"
+	EndpointEmbeddings           = "/v1/embeddings"
+	EndpointAlphaSearch          = "/v1/alpha/search"
+	EndpointResponses            = "/v1/responses"
+	EndpointResponsesCompact     = "/v1/responses/compact"
+	EndpointResponsesInputTokens = "/v1/responses/input_tokens"
+	EndpointImagesGenerations    = "/v1/images/generations"
+	EndpointImagesEdits          = "/v1/images/edits"
+	EndpointImageTasks           = "/v1/images/tasks"
+	EndpointVideosGenerations    = "/v1/videos/generations"
+	EndpointVideosEdits          = "/v1/videos/edits"
+	EndpointVideosExtensions     = "/v1/videos/extensions"
+	EndpointVideos               = "/v1/videos"
+	EndpointGeminiModels         = "/v1beta/models"
 )
 
 const EndpointAntigravityGenerateContent = "/v1internal:streamGenerateContent"
@@ -80,6 +81,8 @@ const (
 func NormalizeInboundEndpoint(path string) string {
 	path = strings.TrimSpace(path)
 	switch {
+	case strings.Contains(path, EndpointResponsesInputTokens) || isResponsesInputTokensAliasPath(path):
+		return EndpointResponsesInputTokens
 	case strings.Contains(path, EndpointEmbeddings):
 		return EndpointEmbeddings
 	case strings.Contains(path, EndpointAlphaSearch) || isBareOrSubpathOf(strings.TrimRight(path, "/"), "/alpha/search") || isBareOrSubpathOf(strings.TrimRight(path, "/"), "/backend-api/codex/alpha/search"):
@@ -113,6 +116,15 @@ func NormalizeInboundEndpoint(path string) string {
 	}
 }
 
+func isResponsesInputTokensAliasPath(path string) bool {
+	trimmed := strings.TrimRight(strings.TrimSpace(path), "/")
+	if trimmed == "" {
+		return false
+	}
+	return isBareOrSubpathOf(trimmed, "/responses/input_tokens") ||
+		isBareOrSubpathOf(trimmed, "/backend-api/codex/responses/input_tokens")
+}
+
 // isResponsesCompactAliasPath reports whether path is the bare/alias
 // "compact" client endpoint — i.e. it is rooted at "/responses/compact"
 // or "/backend-api/codex/responses/compact" (bare routes that serve
@@ -185,7 +197,7 @@ func DeriveUpstreamEndpoint(inbound, rawRequestPath, platform string) string {
 
 	switch platform {
 	case service.PlatformOpenAI, service.PlatformGrok:
-		if inbound == EndpointEmbeddings || inbound == EndpointAlphaSearch || inbound == EndpointImagesGenerations || inbound == EndpointImagesEdits || inbound == EndpointVideosGenerations || inbound == EndpointVideosEdits || inbound == EndpointVideosExtensions || inbound == EndpointVideos {
+		if inbound == EndpointEmbeddings || inbound == EndpointAlphaSearch || inbound == EndpointResponsesInputTokens || inbound == EndpointImagesGenerations || inbound == EndpointImagesEdits || inbound == EndpointVideosGenerations || inbound == EndpointVideosEdits || inbound == EndpointVideosExtensions || inbound == EndpointVideos {
 			return inbound
 		}
 		// OpenAI forwards everything to the Responses API.
@@ -267,7 +279,7 @@ func InboundEndpointMiddleware() gin.HandlerFunc {
 
 // ──────────────────────────────────────────────────────────
 // Context helpers — used by handlers before building
-// RecordUsageInput / RecordUsageLongContextInput.
+// RecordUsageInput.
 // ──────────────────────────────────────────────────────────
 
 // GetInboundEndpoint returns the canonical inbound endpoint stored by
@@ -300,6 +312,14 @@ func GetInboundEndpoint(c *gin.Context) string {
 // and the account platform. Handlers call this after scheduling an
 // account, passing account.Platform.
 func GetUpstreamEndpoint(c *gin.Context, platform string) string {
+	// OpenAI 转发服务维护独立的运行时端点上下文，覆盖普通入站推导。
+	// 这对 force_chat_completions 的错误路径尤为重要：此时可能没有
+	// ForwardResult，不能把入站 /v1/responses 误报成上游端点。
+	if platform == service.PlatformOpenAI || platform == service.PlatformGrok || service.IsCNProvider(platform) {
+		if endpoint := service.GetActualOpenAIUpstreamEndpoint(c); endpoint != "" {
+			return endpoint
+		}
+	}
 	if c != nil {
 		if value, ok := c.Get(ctxKeyActualUpstreamEndpoint); ok {
 			if endpoint, ok := value.(string); ok && endpoint != "" {

backend/internal/handler/endpoint_test.go

@@ -27,6 +27,7 @@ func TestNormalizeInboundEndpoint(t *testing.T) {
 		{"/v1/embeddings", EndpointEmbeddings},
 		{"/v1/alpha/search", EndpointAlphaSearch},
 		{"/v1/responses", EndpointResponses},
+		{"/v1/responses/input_tokens", EndpointResponsesInputTokens},
 		{"/v1/responses/compact", EndpointResponsesCompact},
 		{"/v1/responses/compact/detail", EndpointResponsesCompact},
 		{"/v1/images/generations", EndpointImagesGenerations},
@@ -50,6 +51,7 @@ func TestNormalizeInboundEndpoint(t *testing.T) {
 
 		// Bare top-level alias route "/responses" — root vs. compact.
 		{"/responses", EndpointResponses},
+		{"/responses/input_tokens", EndpointResponsesInputTokens},
 		{"/responses/compact", EndpointResponsesCompact},
 		{"/responses/compact/detail", EndpointResponsesCompact},
 		{"/alpha/search", EndpointAlphaSearch},
@@ -57,6 +59,7 @@ func TestNormalizeInboundEndpoint(t *testing.T) {
 
 		// Bare Codex direct alias route — root vs. compact.
 		{"/backend-api/codex/responses", EndpointResponses},
+		{"/backend-api/codex/responses/input_tokens", EndpointResponsesInputTokens},
 		{"/backend-api/codex/responses/compact", EndpointResponsesCompact},
 		{"/backend-api/codex/responses/compact/detail", EndpointResponsesCompact},
 		{"/backend-api/codex/alpha/search", EndpointAlphaSearch},
@@ -100,6 +103,7 @@ func TestDeriveUpstreamEndpoint(t *testing.T) {
 
 		// OpenAI — root Responses.
 		{"openai responses root", EndpointResponses, "/v1/responses", service.PlatformOpenAI, EndpointResponses},
+		{"openai responses input tokens", EndpointResponsesInputTokens, "/v1/responses/input_tokens", service.PlatformOpenAI, EndpointResponsesInputTokens},
 
 		// OpenAI — compact, raw path carries the derivable "/compact"
 		// (or nested) suffix, which must be preserved on the upstream
@@ -180,6 +184,16 @@ func TestGetUpstreamEndpointPrefersRuntimeOverride(t *testing.T) {
 	require.Equal(t, EndpointMessages, GetUpstreamEndpoint(c, service.PlatformAntigravity))
 }
 
+func TestGetUpstreamEndpointUsesOpenAIRuntimeOverride(t *testing.T) {
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+	c.Request = httptest.NewRequest(http.MethodPost, EndpointResponses, nil)
+	c.Set(ctxKeyInboundEndpoint, EndpointResponses)
+
+	service.SetActualOpenAIUpstreamEndpoint(c, EndpointChatCompletions)
+	require.Equal(t, EndpointChatCompletions, GetUpstreamEndpoint(c, service.PlatformOpenAI))
+}
+
 func TestResolveOpenAIUpstreamEndpointPrefersForwardResult(t *testing.T) {
 	tests := []struct {
 		name            string

backend/internal/handler/failover_loop.go

@@ -56,7 +56,13 @@ const (
 const profitVetoExhaustedMessage = "No available accounts: all candidates rejected by group profit control"
 
 func sameAccountRetryDelayFor(failoverErr *service.UpstreamFailoverError, retryCount int) time.Duration {
-	if failoverErr == nil || !failoverErr.RequestScopedTransient || retryCount <= 1 {
+	if failoverErr == nil {
+		return sameAccountRetryDelay
+	}
+	if failoverErr.SameAccountRetryDelay > 0 {
+		return failoverErr.SameAccountRetryDelay
+	}
+	if !failoverErr.RequestScopedTransient || retryCount <= 1 {
 		return sameAccountRetryDelay
 	}
 
@@ -70,6 +76,51 @@ func sameAccountRetryDelayFor(failoverErr *service.UpstreamFailoverError, retryC
 	return delay
 }
 
+func sameAccountRetryAllowed(failoverErr *service.UpstreamFailoverError, retryCount, retryLimit int) bool {
+	if failoverErr == nil || !failoverErr.RetryableOnSameAccount {
+		return false
+	}
+	if !sameAccountRetryDeadlineAllows(failoverErr) {
+		return false
+	}
+	// Error-specific caps (Grok capacity/stream-idle) remain hard limits even
+	// when the error also carries a freshly reconstructed deadline.
+	if failoverErr.SameAccountRetryMax > 0 {
+		if retryLimit <= 0 {
+			return false
+		}
+		if failoverErr.SameAccountRetryMax < retryLimit {
+			retryLimit = failoverErr.SameAccountRetryMax
+		}
+		return retryCount < retryLimit
+	}
+	// OAuth 429 explicitly opts into a deadline window. It is intentionally not
+	// bounded by the ordinary/default pool retry count.
+	if !failoverErr.SameAccountRetryDeadline.IsZero() {
+		return true
+	}
+	return retryLimit > 0 && retryCount < retryLimit
+}
+
+// sameAccountRetryDeadlineAllows prevents a retry from starting after the
+// service-provided same-account retry window has elapsed.
+func sameAccountRetryDeadlineAllows(failoverErr *service.UpstreamFailoverError) bool {
+	return failoverErr == nil || failoverErr.SameAccountRetryDeadline.IsZero() || time.Now().Before(failoverErr.SameAccountRetryDeadline)
+}
+
+// effectiveSameAccountRetryLimit applies an error-specific cap without
+// overriding an explicit account setting of zero (which disables retries).
+func effectiveSameAccountRetryLimit(failoverErr *service.UpstreamFailoverError, account *service.Account) int {
+	if account == nil {
+		return 0
+	}
+	limit := account.GetPoolModeRetryCount()
+	if limit > 0 && failoverErr != nil && failoverErr.SameAccountRetryMax > 0 && failoverErr.SameAccountRetryMax < limit {
+		return failoverErr.SameAccountRetryMax
+	}
+	return limit
+}
+
 // FailoverState 跨循环迭代共享的 failover 状态
 type FailoverState struct {
 	SwitchCount           int
@@ -158,14 +209,15 @@ func (s *FailoverState) HandleFailoverError(
 	}
 
 	// 同账号重试不算切换账号，粘性会话仅在实际切换时强制缓存计费。
-	sameAccountRetry := failoverErr.RetryableOnSameAccount && s.SameAccountRetryCount[accountID] < retryLimit
+	retryCount := s.SameAccountRetryCount[accountID]
+	sameAccountRetry := sameAccountRetryAllowed(failoverErr, retryCount, retryLimit)
 	if needForceCacheBilling(s.hasBoundSession, failoverErr, sameAccountRetry) {
 		s.ForceCacheBilling = true
 	}
 
 	// 同账号重试：对 RetryableOnSameAccount 的临时性错误，先在同一账号上重试。
 	// 重试次数上限 retryLimit 由调用方传入（账号级 pool_mode_retry_count 配置）。
-	if failoverErr.RetryableOnSameAccount && s.SameAccountRetryCount[accountID] < retryLimit {
+	if sameAccountRetry {
 		s.SameAccountRetryCount[accountID]++
 		retryDelay := sameAccountRetryDelayFor(failoverErr, s.SameAccountRetryCount[accountID])
 		logger.FromContext(ctx).Warn("gateway.failover_same_account_retry",

backend/internal/handler/failover_loop_test.go

@@ -58,6 +58,60 @@ func TestSameAccountRetryDelayFor(t *testing.T) {
 	t.Run("nil error keeps fixed delay", func(t *testing.T) {
 		require.Equal(t, 500*time.Millisecond, sameAccountRetryDelayFor(nil, 10))
 	})
+
+	t.Run("explicit oauth delay wins", func(t *testing.T) {
+		err := &service.UpstreamFailoverError{SameAccountRetryDelay: 3 * time.Second}
+		require.Equal(t, 3*time.Second, sameAccountRetryDelayFor(err, 1))
+	})
+}
+
+func TestSameAccountRetryAllowedUsesDeadlineInsteadOfPoolCount(t *testing.T) {
+	err := &service.UpstreamFailoverError{
+		RetryableOnSameAccount:   true,
+		SameAccountRetryDeadline: time.Now().Add(time.Minute),
+	}
+	require.True(t, sameAccountRetryAllowed(err, 100, 0))
+	require.True(t, sameAccountRetryAllowed(err, 100, maxSameAccountRetries))
+	err.SameAccountRetryDeadline = time.Now().Add(-time.Second)
+	require.False(t, sameAccountRetryAllowed(err, 0, 100))
+}
+
+func TestSameAccountRetryAllowedRequiresOptInAndDefaultsToCountLimit(t *testing.T) {
+	err := &service.UpstreamFailoverError{SameAccountRetryDeadline: time.Now().Add(time.Minute)}
+	require.False(t, sameAccountRetryAllowed(err, 0, maxSameAccountRetries))
+
+	err.RetryableOnSameAccount = true
+	err.SameAccountRetryDeadline = time.Time{}
+	require.True(t, sameAccountRetryAllowed(err, maxSameAccountRetries-1, maxSameAccountRetries))
+	require.False(t, sameAccountRetryAllowed(err, maxSameAccountRetries, maxSameAccountRetries))
+}
+
+func TestSameAccountRetryAllowedHonorsErrorMaxBeforeDeadline(t *testing.T) {
+	err := &service.UpstreamFailoverError{
+		RetryableOnSameAccount:   true,
+		SameAccountRetryDeadline: time.Now().Add(time.Minute),
+		SameAccountRetryMax:      1,
+	}
+	require.True(t, sameAccountRetryAllowed(err, 0, maxSameAccountRetries))
+	require.False(t, sameAccountRetryAllowed(err, 1, maxSameAccountRetries))
+	require.False(t, sameAccountRetryAllowed(err, 0, 0), "an explicit zero retry budget remains disabled")
+}
+
+func TestSameAccountRetryDeadlineAllows(t *testing.T) {
+	require.True(t, sameAccountRetryDeadlineAllows(&service.UpstreamFailoverError{}))
+	require.True(t, sameAccountRetryDeadlineAllows(&service.UpstreamFailoverError{
+		SameAccountRetryDeadline: time.Now().Add(time.Second),
+	}))
+	require.False(t, sameAccountRetryDeadlineAllows(&service.UpstreamFailoverError{
+		SameAccountRetryDeadline: time.Now().Add(-time.Second),
+	}))
+}
+
+func TestEffectiveSameAccountRetryLimitHonorsErrorCapAndDisabledAccount(t *testing.T) {
+	account := &service.Account{Type: service.AccountTypeAPIKey, Credentials: map[string]any{"pool_mode": true, "pool_mode_retry_count": float64(3)}}
+	require.Equal(t, 1, effectiveSameAccountRetryLimit(&service.UpstreamFailoverError{SameAccountRetryMax: 1}, account))
+	account.Credentials["pool_mode_retry_count"] = float64(0)
+	require.Equal(t, 0, effectiveSameAccountRetryLimit(&service.UpstreamFailoverError{SameAccountRetryMax: 1}, account))
 }
 
 // ---------------------------------------------------------------------------
@@ -320,6 +374,22 @@ func TestHandleFailoverError_CacheBilling(t *testing.T) {
 		require.Zero(t, fs.SwitchCount)
 	})
 
+	t.Run("OAuth deadline存在时不按普通计数切换", func(t *testing.T) {
+		mock := &mockTempUnscheduler{}
+		fs := NewFailoverState(3, true)
+		fs.SameAccountRetryCount[100] = maxSameAccountRetries
+		err := newTestFailoverErr(http.StatusTooManyRequests, true, false)
+		err.SameAccountRetryDeadline = time.Now().Add(time.Minute)
+		err.SameAccountRetryDelay = time.Nanosecond
+
+		fs.HandleFailoverError(context.Background(), mock, 100, "openai", maxSameAccountRetries, err)
+
+		require.False(t, fs.ForceCacheBilling)
+		require.Zero(t, fs.SwitchCount)
+		require.Equal(t, maxSameAccountRetries+1, fs.SameAccountRetryCount[100])
+		require.Empty(t, mock.calls)
+	})
+
 	t.Run("同账号重试耗尽并实际切换时设置ForceCacheBilling", func(t *testing.T) {
 		mock := &mockTempUnscheduler{}
 		fs := NewFailoverState(3, true)

backend/internal/handler/gateway_handler.go

@@ -169,6 +169,7 @@ func (h *GatewayHandler) Messages(c *gin.Context) {
 	body = parsedReq.Body.Bytes()
 	reqModel := parsedReq.Model
 	reqStream := parsedReq.Stream
+	bindRequestedReasoningEffort(c, body, reqModel)
 	ensureCompositeTargetPlatform(c, apiKey, reqModel)
 	reqLog = reqLog.With(zap.String("model", reqModel), zap.Bool("stream", reqStream))
 
@@ -541,6 +542,7 @@ func (h *GatewayHandler) Messages(c *gin.Context) {
 			inboundEndpoint := GetInboundEndpoint(c)
 			upstreamEndpoint := GetUpstreamEndpoint(c, account.Platform)
 
+			stampForwardRequestedReasoningEffort(result, service.NormalizeClaudeOutputEffort(parsedReq.OutputEffort))
 			if result.ReasoningEffort == nil {
 				result.ReasoningEffort = service.NormalizeClaudeOutputEffort(parsedReq.OutputEffort)
 			}
@@ -882,6 +884,7 @@ func (h *GatewayHandler) Messages(c *gin.Context) {
 				inboundEndpoint := GetInboundEndpoint(c)
 				upstreamEndpoint := GetUpstreamEndpoint(c, account.Platform)
 
+				stampForwardRequestedReasoningEffort(result, service.NormalizeClaudeOutputEffort(attemptParsedReq.OutputEffort))
 				if result.ReasoningEffort == nil {
 					result.ReasoningEffort = service.NormalizeClaudeOutputEffort(attemptParsedReq.OutputEffort)
 				}
@@ -1140,6 +1143,79 @@ func (h *GatewayHandler) Models(c *gin.Context) {
 	})
 }
 
+// CodexModels returns the effective group model list using the manifest shape
+// expected by Codex custom providers. Official OpenAI groups continue to use
+// OpenAIGatewayHandler.CodexModels so their live upstream metadata is preserved.
+func (h *GatewayHandler) CodexModels(c *gin.Context) {
+	apiKey, ok := middleware2.GetAPIKeyFromContext(c)
+	if !ok || apiKey == nil || apiKey.Group == nil {
+		h.errorResponse(c, http.StatusUnauthorized, "invalid_request_error", "API key group is required")
+		return
+	}
+
+	forcedPlatform := ""
+	if value, exists := middleware2.GetForcePlatformFromContext(c); exists {
+		forcedPlatform = strings.TrimSpace(value)
+	}
+	modelIDs := h.codexModelIDsForGroup(c.Request.Context(), apiKey.Group, forcedPlatform)
+	modelIDs = service.FilterCodexModelIDsForGroup(modelIDs, apiKey.Group)
+	body, err := h.gatewayService.BuildCodexModelsManifestForGroup(
+		c.Request.Context(),
+		apiKey.Group,
+		forcedPlatform,
+		modelIDs,
+	)
+	if err != nil {
+		h.errorResponse(c, http.StatusInternalServerError, "api_error", "Failed to build Codex models manifest")
+		return
+	}
+	etag := service.CodexModelsManifestETag(body)
+	c.Header("ETag", etag)
+	if service.CodexModelsManifestETagMatches(c.GetHeader("If-None-Match"), etag) {
+		c.Status(http.StatusNotModified)
+		c.Writer.WriteHeaderNow()
+		return
+	}
+	c.Data(http.StatusOK, "application/json", body)
+}
+
+func (h *GatewayHandler) codexModelIDsForGroup(ctx context.Context, group *service.Group, platformOverride string) []string {
+	if h == nil || h.gatewayService == nil || group == nil {
+		return nil
+	}
+
+	groupID := &group.ID
+	platform := strings.TrimSpace(platformOverride)
+	if platform == "" {
+		platform = group.Platform
+	}
+	if platform == service.PlatformComposite {
+		availableModels := h.compositeAvailableModels(ctx, groupID)
+		fallbackModels := defaultCodexModelIDsForPlatform(service.PlatformComposite)
+		if group.CustomModelsListEnabled() {
+			return filterModelsByCustomList(availableModels, fallbackModels, group.ModelsListConfig.Models)
+		}
+		if len(availableModels) > 0 {
+			return availableModels
+		}
+		return fallbackModels
+	}
+
+	availableModels := h.gatewayService.GetAvailableModels(ctx, groupID, platform)
+	fallbackModels := defaultCodexModelIDsForPlatform(platform)
+	if group.CustomModelsListEnabled() {
+		return filterModelsByCustomList(
+			customModelsListSource(platform, availableModels, fallbackModels),
+			fallbackModels,
+			group.ModelsListConfig.Models,
+		)
+	}
+	if len(availableModels) > 0 {
+		return availableModels
+	}
+	return fallbackModels
+}
+
 func (h *GatewayHandler) compositeAvailableModels(ctx context.Context, groupID *int64) []string {
 	if h == nil || h.gatewayService == nil {
 		return nil
@@ -1147,10 +1223,12 @@ func (h *GatewayHandler) compositeAvailableModels(ctx context.Context, groupID *
 	seen := make(map[string]struct{})
 	models := make([]string, 0)
 	schedulablePlatforms := h.gatewayService.GetSchedulablePlatforms(ctx, groupID)
-	for _, platform := range []string{service.PlatformAnthropic, service.PlatformGemini, service.PlatformOpenAI, service.PlatformAntigravity, service.PlatformGrok} {
+	for _, platform := range []string{service.PlatformAnthropic, service.PlatformGemini, service.PlatformOpenAI, service.PlatformAntigravity, service.PlatformGrok, service.PlatformKimi, service.PlatformZhipu, service.PlatformDeepseek} {
 		platformModels := h.gatewayService.GetAvailableModels(ctx, groupID, platform)
 		if len(platformModels) == 0 {
-			if _, ok := schedulablePlatforms[platform]; ok {
+			// CN 供应商没有静态默认模型列表（defaultModelIDsForPlatform 的
+			// default 分支是 Claude 列表），composite 下只暴露账号映射键。
+			if _, ok := schedulablePlatforms[platform]; ok && !service.IsCNProvider(platform) {
 				platformModels = defaultModelIDsForPlatform(platform)
 			}
 		}
@@ -1232,11 +1310,15 @@ func writeGrokModelsList(c *gin.Context, modelIDs []string) {
 		if grokModelSupportsConfigurableReasoning(modelID) {
 			item.SupportsReasoningEffort = true
 			item.ReasoningEffort = "high"
-			item.ReasoningEfforts = []grokReasoningEffortOption{
+			efforts := []grokReasoningEffortOption{
 				{Value: "low", Label: "Low"},
 				{Value: "medium", Label: "Medium"},
 				{Value: "high", Label: "High", Default: true},
 			}
+			if service.GrokSupportsXHighReasoningEffort(modelID) {
+				efforts = append(efforts, grokReasoningEffortOption{Value: "xhigh", Label: "xHigh"})
+			}
+			item.ReasoningEfforts = efforts
 		}
 		models = append(models, item)
 	}
@@ -1338,9 +1420,26 @@ func customModelsListAllowsModel(availablePatterns []string, model string) bool
 			return true
 		}
 	}
+	normalizedClaudeModel := claude.NormalizeModelID(strings.TrimSuffix(model, "-thinking"))
+	if normalizedClaudeModel != model {
+		for _, pattern := range availablePatterns {
+			if pattern == normalizedClaudeModel {
+				return true
+			}
+		}
+	}
 	return false
 }
 
+func defaultCodexModelIDsForPlatform(platform string) []string {
+	switch platform {
+	case service.PlatformDeepseek:
+		return []string{"deepseek-v4-pro", "deepseek-v4-flash"}
+	default:
+		return defaultModelIDsForPlatform(platform)
+	}
+}
+
 func defaultModelIDsForPlatform(platform string) []string {
 	switch platform {
 	case service.PlatformOpenAI:
@@ -1359,20 +1458,13 @@ func defaultModelIDsForPlatform(platform string) []string {
 		}
 		return ids
 	case service.PlatformAnthropic:
-		ids := make([]string, 0, len(claude.DefaultModels)+len(antigravity.DefaultModels()))
-		for _, model := range claude.DefaultModels {
-			ids = append(ids, model.ID)
-		}
-		for _, model := range antigravity.DefaultModels() {
-			ids = append(ids, model.ID)
-		}
-		return mergeModelIDs(ids, nil)
+		return claude.DefaultModelIDs()
 	case service.PlatformGrok:
 		return xai.DefaultModelIDs()
 	case service.PlatformComposite:
 		ids := make([]string, 0)
 		seen := make(map[string]struct{})
-		for _, concretePlatform := range []string{service.PlatformAnthropic, service.PlatformGemini, service.PlatformOpenAI, service.PlatformAntigravity, service.PlatformGrok} {
+		for _, concretePlatform := range []string{service.PlatformAnthropic, service.PlatformGemini, service.PlatformOpenAI, service.PlatformAntigravity, service.PlatformGrok, service.PlatformKimi, service.PlatformZhipu, service.PlatformDeepseek} {
 			for _, id := range defaultModelIDsForPlatform(concretePlatform) {
 				if _, ok := seen[id]; ok {
 					continue

backend/internal/handler/gateway_handler_chat_completions.go

@@ -5,6 +5,7 @@ import (
 	"errors"
 	"net/http"
 	"strconv"
+	"strings"
 	"time"
 
 	"github.com/Wei-Shaw/sub2api/internal/pkg/ip"
@@ -75,6 +76,7 @@ func (h *GatewayHandler) ChatCompletions(c *gin.Context) {
 		return
 	}
 	reqModel := modelResult.String()
+	bindRequestedReasoningEffort(c, body, reqModel)
 	ensureCompositeTargetPlatform(c, apiKey, reqModel)
 	if !compositeTargetPlatformResolved(c, apiKey, reqModel) {
 		h.chatCompletionsErrorResponse(c, http.StatusBadRequest, "invalid_request_error", "Model is not supported by composite groups")
@@ -173,6 +175,7 @@ func (h *GatewayHandler) ChatCompletions(c *gin.Context) {
 		if err != nil {
 			if len(fs.FailedAccountIDs) == 0 {
 				cls := classifyNoAccountErrorFromGin(c, h.gatewayService, apiKey, reqModel, reqModel, groupPlatform)
+				cls = classifySelectionFailureError(err, cls)
 				if !cls.ModelNotFound {
 					markOpsRoutingCapacityLimitedIfNoAvailable(c, err)
 				}
@@ -333,6 +336,7 @@ func (h *GatewayHandler) ChatCompletions(c *gin.Context) {
 
 		quotaPlatform := service.QuotaPlatform(c.Request.Context(), apiKey)
 		sessionID := service.ExtractClientSessionID(c)
+		stampForwardRequestedReasoningEffort(result, service.RequestedReasoningEffortFromContext(c.Request.Context()))
 		h.submitUsageRecordTask(c.Request.Context(), func(ctx context.Context) {
 			if err := h.gatewayService.RecordUsage(ctx, &service.RecordUsageInput{
 				Result:             result,
@@ -384,6 +388,14 @@ func (h *GatewayHandler) handleCCFailoverExhausted(c *gin.Context, lastErr *serv
 		h.chatCompletionsErrorResponse(c, status, "server_error", message)
 		return
 	}
+	if lastErr != nil && lastErr.IsOpenAICapacityShed() && strings.TrimSpace(lastErr.ClientMessage) != "" {
+		status := lastErr.ClientStatusCode
+		if status <= 0 {
+			status = http.StatusServiceUnavailable
+		}
+		h.chatCompletionsErrorResponse(c, status, "server_error", lastErr.ClientMessage)
+		return
+	}
 	statusCode := http.StatusBadGateway
 	if lastErr != nil && lastErr.StatusCode > 0 {
 		statusCode = lastErr.StatusCode

backend/internal/handler/gateway_handler_responses.go

@@ -5,6 +5,7 @@ import (
 	"errors"
 	"net/http"
 	"strconv"
+	"strings"
 	"time"
 
 	"github.com/Wei-Shaw/sub2api/internal/pkg/ip"
@@ -75,6 +76,7 @@ func (h *GatewayHandler) Responses(c *gin.Context) {
 		return
 	}
 	reqModel := modelResult.String()
+	bindRequestedReasoningEffort(c, body, reqModel)
 	ensureCompositeTargetPlatform(c, apiKey, reqModel)
 	if !compositeTargetPlatformResolved(c, apiKey, reqModel) {
 		h.responsesErrorResponse(c, http.StatusBadRequest, "invalid_request_error", "Model is not supported by composite groups")
@@ -175,6 +177,7 @@ func (h *GatewayHandler) Responses(c *gin.Context) {
 		if err != nil {
 			if len(fs.FailedAccountIDs) == 0 {
 				cls := classifyNoAccountErrorFromGin(c, h.gatewayService, apiKey, reqModel, reqModel, effectiveAPIKeyPlatform(c, apiKey))
+				cls = classifySelectionFailureError(err, cls)
 				if !cls.ModelNotFound {
 					markOpsRoutingCapacityLimitedIfNoAvailable(c, err)
 				}
@@ -321,6 +324,7 @@ func (h *GatewayHandler) Responses(c *gin.Context) {
 
 		quotaPlatform := service.QuotaPlatform(c.Request.Context(), apiKey)
 		sessionID := service.ExtractClientSessionID(c)
+		stampForwardRequestedReasoningEffort(result, service.RequestedReasoningEffortFromContext(c.Request.Context()))
 		h.submitUsageRecordTask(c.Request.Context(), func(ctx context.Context) {
 			if err := h.gatewayService.RecordUsage(ctx, &service.RecordUsageInput{
 				Result:             result,
@@ -361,25 +365,38 @@ func (h *GatewayHandler) responsesErrorResponse(c *gin.Context, status int, code
 
 // handleResponsesFailoverExhausted writes a failover-exhausted error in Responses format.
 func (h *GatewayHandler) handleResponsesFailoverExhausted(c *gin.Context, lastErr *service.UpstreamFailoverError, streamStarted bool) {
-	if streamStarted {
-		return // Can't write error after stream started
-	}
 	if lastErr != nil {
 		copyFailoverRetryAfter(c, lastErr.ResponseHeaders)
 	}
-	if lastErr != nil && lastErr.IsCredentialFailure() {
-		status, message := credentialFailoverClientResponse(lastErr)
-		h.responsesErrorResponse(c, status, "server_error", message)
-		return
-	}
 	statusCode := http.StatusBadGateway
 	if lastErr != nil && lastErr.StatusCode > 0 {
 		statusCode = lastErr.StatusCode
 	}
-	if lastErr != nil && service.IsOpenAISilentRefusalErrorBody(lastErr.ResponseBody) {
+	status, code, message := statusCode, "server_error", "All available accounts exhausted"
+	if lastErr != nil && lastErr.IsCredentialFailure() {
+		status, message = credentialFailoverClientResponse(lastErr)
+	} else if lastErr != nil && lastErr.IsOpenAICapacityShed() && strings.TrimSpace(lastErr.ClientMessage) != "" {
+		status = lastErr.ClientStatusCode
+		if status <= 0 {
+			status = http.StatusServiceUnavailable
+		}
+		message = lastErr.ClientMessage
+	} else if lastErr != nil && service.IsOpenAISilentRefusalErrorBody(lastErr.ResponseBody) {
 		service.SetOpsUpstreamError(c, statusCode, service.OpenAISilentRefusalClientMessage(), "")
-		h.responsesErrorResponse(c, http.StatusBadGateway, "upstream_error", service.OpenAISilentRefusalClientMessage())
+		status, code, message = http.StatusBadGateway, "upstream_error", service.OpenAISilentRefusalClientMessage()
+	} else if lastErr != nil && statusCode == http.StatusTooManyRequests {
+		status, code, message = http.StatusTooManyRequests, "rate_limit_error", "All available accounts are currently rate-limited. Please retry later."
+	}
+	if streamStarted {
+		// A slot-wait heartbeat commits HTTP 200 before any upstream response.
+		// In that case a terminal frame is still required; once any semantic or
+		// official terminal bytes exist, preserve them without appending a second
+		// generic response.failed.
+		service.MarkOpsStreamError(c, code, message, status)
+		if c != nil && c.Writer != nil && (c.Writer.Size() <= 0 || gatewayStreamHasOnlyHeartbeats(c)) {
+			writeResponsesFailedSSE(c, code, message)
+		}
 		return
 	}
-	h.responsesErrorResponse(c, statusCode, "server_error", "All available accounts exhausted")
+	h.responsesErrorResponse(c, status, code, message)
 }

backend/internal/handler/gateway_helper.go

@@ -16,6 +16,29 @@ import (
 	"github.com/gin-gonic/gin"
 )
 
+const gatewayStreamHeartbeatBytesKey = "gateway_stream_heartbeat_bytes"
+
+func recordGatewayStreamHeartbeat(c *gin.Context, written int) {
+	if c == nil || written <= 0 {
+		return
+	}
+	total, _ := c.Get(gatewayStreamHeartbeatBytesKey)
+	bytes, _ := total.(int)
+	c.Set(gatewayStreamHeartbeatBytesKey, bytes+written)
+}
+
+func gatewayStreamHasOnlyHeartbeats(c *gin.Context) bool {
+	if c == nil || c.Writer == nil {
+		return false
+	}
+	value, ok := c.Get(gatewayStreamHeartbeatBytesKey)
+	if !ok {
+		return false
+	}
+	heartbeatBytes, _ := value.(int)
+	return heartbeatBytes > 0 && c.Writer.Size() == heartbeatBytes
+}
+
 // claudeCodeValidator is a singleton validator for Claude Code client detection
 var claudeCodeValidator = service.NewClaudeCodeValidator()
 
@@ -396,9 +419,11 @@ func (h *ConcurrencyHelper) waitForSlotWithPingTimeout(c *gin.Context, slotType
 				c.Header("X-Accel-Buffering", "no")
 				*streamStarted = true
 			}
-			if _, err := fmt.Fprint(c.Writer, string(h.pingFormat)); err != nil {
+			written, err := fmt.Fprint(c.Writer, string(h.pingFormat))
+			if err != nil {
 				return nil, err
 			}
+			recordGatewayStreamHeartbeat(c, written)
 			flusher.Flush()
 
 		case <-timer.C:

backend/internal/handler/gateway_models_test.go

@@ -7,6 +7,7 @@ import (
 	"net/http/httptest"
 	"testing"
 
+	"github.com/Wei-Shaw/sub2api/internal/pkg/claude"
 	middleware2 "github.com/Wei-Shaw/sub2api/internal/server/middleware"
 	"github.com/Wei-Shaw/sub2api/internal/service"
 	"github.com/gin-gonic/gin"
@@ -24,6 +25,22 @@ type gatewayModelsResponseForTest struct {
 	Data   []gatewayModelItemForTest `json:"data"`
 }
 
+type codexModelsResponseForTest struct {
+	Models []struct {
+		Slug                     string                       `json:"slug"`
+		SupportedReasoningLevels []codexReasoningLevelForTest `json:"supported_reasoning_levels"`
+		InputModalities          []string                     `json:"input_modalities"`
+		ModelMessages            map[string]json.RawMessage   `json:"model_messages"`
+		TruncationPolicy         map[string]json.RawMessage   `json:"truncation_policy"`
+		AvailabilityNUX          json.RawMessage              `json:"availability_nux"`
+		Upgrade                  json.RawMessage              `json:"upgrade"`
+	} `json:"models"`
+}
+
+type codexReasoningLevelForTest struct {
+	Effort string `json:"effort"`
+}
+
 type gatewayModelItemForTest struct {
 	ID                      string                                `json:"id"`
 	Object                  string                                `json:"object"`
@@ -51,6 +68,17 @@ func (s *gatewayModelsAccountRepoStub) ListSchedulableByGroupID(ctx context.Cont
 	return out, nil
 }
 
+func (s *gatewayModelsAccountRepoStub) ListByGroup(ctx context.Context, groupID int64) ([]service.Account, error) {
+	return s.ListSchedulableByGroupID(ctx, groupID)
+}
+
+func (s *gatewayModelsAccountRepoStub) ListModelAvailabilityCandidates(ctx context.Context, groupID *int64, _ []string, _ bool) ([]service.Account, error) {
+	if groupID == nil {
+		return nil, nil
+	}
+	return s.ListSchedulableByGroupID(ctx, *groupID)
+}
+
 func newGatewayModelsHandlerForTest(repo service.AccountRepository) *GatewayHandler {
 	return &GatewayHandler{
 		gatewayService: service.NewGatewayService(
@@ -69,6 +97,247 @@ func TestDefaultModelIDsForCompositeIncludesAntigravityDefaults(t *testing.T) {
 	require.Contains(t, compositeIDs, antigravityIDs[0])
 }
 
+// Scenario: Anthropic defaults contain only Claude while Antigravity keeps its own Gemini models.
+func TestDefaultModelIDsForAnthropicExcludeAntigravityGemini(t *testing.T) {
+	anthropicIDs := defaultModelIDsForPlatform(service.PlatformAnthropic)
+	require.Contains(t, anthropicIDs, "claude-opus-4-6")
+	require.NotContains(t, anthropicIDs, "gemini-2.5-flash")
+
+	antigravityIDs := defaultModelIDsForPlatform(service.PlatformAntigravity)
+	require.Contains(t, antigravityIDs, "gemini-2.5-flash")
+}
+
+// Scenario: non-OpenAI groups return a Codex manifest instead of a standard model list.
+func TestGatewayCodexModels_NonOpenAIGroupsUseMappedModels(t *testing.T) {
+	tests := []struct {
+		name       string
+		platform   string
+		model      string
+		efforts    []string
+		modalities []string
+	}{
+		{
+			name:       "Grok",
+			platform:   service.PlatformGrok,
+			model:      "grok-4.6",
+			efforts:    []string{"low", "medium", "high", "xhigh"},
+			modalities: []string{"text", "image"},
+		},
+		{
+			name:       "DeepSeek",
+			platform:   service.PlatformDeepseek,
+			model:      "deepseek-v4-pro",
+			efforts:    []string{"low", "high", "max"},
+			modalities: []string{"text"},
+		},
+		{
+			name:       "provider-qualified Claude",
+			platform:   service.PlatformAnthropic,
+			model:      "anthropic/claude-sonnet-4-6",
+			efforts:    []string{"low", "medium", "high", "max"},
+			modalities: []string{"text"},
+		},
+	}
+
+	for index, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			gin.SetMode(gin.TestMode)
+			groupID := int64(100 + index)
+			h := newGatewayModelsHandlerForTest(&gatewayModelsAccountRepoStub{
+				byGroup: map[int64][]service.Account{
+					groupID: {
+						{
+							ID:       1,
+							Platform: tt.platform,
+							Credentials: map[string]any{
+								"model_mapping": map[string]any{tt.model: tt.model},
+							},
+						},
+					},
+				},
+			})
+
+			rec := httptest.NewRecorder()
+			c, _ := gin.CreateTestContext(rec)
+			c.Request = httptest.NewRequest(http.MethodGet, "/models?client_version=0.147.0", nil)
+			c.Set(string(middleware2.ContextKeyAPIKey), &service.APIKey{
+				Group: &service.Group{ID: groupID, Platform: tt.platform},
+			})
+
+			h.CodexModels(c)
+
+			require.Equal(t, http.StatusOK, rec.Code)
+			var got codexModelsResponseForTest
+			require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &got))
+			require.Len(t, got.Models, 1)
+			require.Equal(t, tt.model, got.Models[0].Slug)
+			require.NotEmpty(t, got.Models[0].ModelMessages)
+			require.NotEmpty(t, got.Models[0].TruncationPolicy)
+			require.NotNil(t, got.Models[0].AvailabilityNUX)
+			require.NotNil(t, got.Models[0].Upgrade)
+			require.Equal(t, tt.efforts, codexReasoningEffortsForTest(got.Models[0].SupportedReasoningLevels))
+			require.Equal(t, tt.modalities, got.Models[0].InputModalities)
+		})
+	}
+}
+
+// Scenario: Composite manifests aggregate only administrator-configured models.
+func TestGatewayCodexModels_CompositeUsesCompleteEffectiveModelList(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	const groupID int64 = 120
+	h := newGatewayModelsHandlerForTest(&gatewayModelsAccountRepoStub{
+		byGroup: map[int64][]service.Account{
+			groupID: {
+				{
+					ID:          3,
+					Platform:    service.PlatformOpenAI,
+					Status:      service.StatusActive,
+					Schedulable: true,
+					Credentials: map[string]any{},
+				},
+				{
+					ID:       1,
+					Platform: service.PlatformOpenAI,
+					Credentials: map[string]any{
+						"model_mapping": map[string]any{"gpt-5.5": "gpt-5.5"},
+					},
+				},
+				{
+					ID:       2,
+					Platform: service.PlatformGrok,
+					Credentials: map[string]any{
+						"model_mapping": map[string]any{"grok-4.6": "grok-4.6"},
+					},
+				},
+			},
+		},
+	})
+
+	rec := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(rec)
+	c.Request = httptest.NewRequest(http.MethodGet, "/models?client_version=0.147.0", nil)
+	c.Set(string(middleware2.ContextKeyAPIKey), &service.APIKey{
+		Group: &service.Group{ID: groupID, Platform: service.PlatformComposite},
+	})
+
+	h.CodexModels(c)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	var got codexModelsResponseForTest
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &got))
+	require.Equal(t, []string{"gpt-5.5", "grok-4.6"}, codexModelSlugsForTest(got.Models))
+}
+
+func TestGatewayCodexModels_GeneratedManifestUsesFinalBodyETag(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	const groupID int64 = 122
+	h := newGatewayModelsHandlerForTest(&gatewayModelsAccountRepoStub{
+		byGroup: map[int64][]service.Account{
+			groupID: {{
+				ID:       1,
+				Platform: service.PlatformDeepseek,
+				Credentials: map[string]any{
+					"model_mapping": map[string]any{"deepseek-v4-pro": "deepseek-v4-pro"},
+				},
+			}},
+		},
+	})
+	group := &service.Group{ID: groupID, Platform: service.PlatformDeepseek}
+
+	first := httptest.NewRecorder()
+	firstContext, _ := gin.CreateTestContext(first)
+	firstContext.Request = httptest.NewRequest(http.MethodGet, "/models?client_version=0.147.0", nil)
+	firstContext.Set(string(middleware2.ContextKeyAPIKey), &service.APIKey{Group: group})
+	h.CodexModels(firstContext)
+
+	require.Equal(t, http.StatusOK, first.Code)
+	etag := first.Header().Get("ETag")
+	require.NotEmpty(t, etag)
+	require.Equal(t, service.CodexModelsManifestETag(first.Body.Bytes()), etag)
+
+	second := httptest.NewRecorder()
+	secondContext, _ := gin.CreateTestContext(second)
+	secondContext.Request = httptest.NewRequest(http.MethodGet, "/models?client_version=0.147.0", nil)
+	secondContext.Request.Header.Set("If-None-Match", "W/"+etag)
+	secondContext.Set(string(middleware2.ContextKeyAPIKey), &service.APIKey{Group: group})
+	h.CodexModels(secondContext)
+
+	require.Equal(t, http.StatusNotModified, second.Code)
+	require.Empty(t, second.Body.Bytes())
+	require.Equal(t, etag, second.Header().Get("ETag"))
+}
+
+// Scenario: group models_list_config limits the generated Codex manifest.
+func TestGatewayCodexModels_CustomModelsListFiltersCompositeManifest(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	const groupID int64 = 121
+	h := newGatewayModelsHandlerForTest(&gatewayModelsAccountRepoStub{
+		byGroup: map[int64][]service.Account{
+			groupID: {
+				{
+					ID:       1,
+					Platform: service.PlatformOpenAI,
+					Credentials: map[string]any{
+						"model_mapping": map[string]any{"gpt-5.5": "gpt-5.5"},
+					},
+				},
+				{
+					ID:       2,
+					Platform: service.PlatformGrok,
+					Credentials: map[string]any{
+						"model_mapping": map[string]any{"grok-4.6": "grok-4.6"},
+					},
+				},
+			},
+		},
+	})
+
+	rec := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(rec)
+	c.Request = httptest.NewRequest(http.MethodGet, "/models?client_version=0.147.0", nil)
+	c.Set(string(middleware2.ContextKeyAPIKey), &service.APIKey{
+		Group: &service.Group{
+			ID:       groupID,
+			Platform: service.PlatformComposite,
+			ModelsListConfig: service.GroupModelsListConfig{
+				Enabled: true,
+				Models:  []string{"grok-4.6"},
+			},
+		},
+	})
+
+	h.CodexModels(c)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	var got codexModelsResponseForTest
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &got))
+	require.Equal(t, []string{"grok-4.6"}, codexModelSlugsForTest(got.Models))
+}
+
+func codexModelSlugsForTest(models []struct {
+	Slug                     string                       `json:"slug"`
+	SupportedReasoningLevels []codexReasoningLevelForTest `json:"supported_reasoning_levels"`
+	InputModalities          []string                     `json:"input_modalities"`
+	ModelMessages            map[string]json.RawMessage   `json:"model_messages"`
+	TruncationPolicy         map[string]json.RawMessage   `json:"truncation_policy"`
+	AvailabilityNUX          json.RawMessage              `json:"availability_nux"`
+	Upgrade                  json.RawMessage              `json:"upgrade"`
+}) []string {
+	slugs := make([]string, 0, len(models))
+	for _, model := range models {
+		slugs = append(slugs, model.Slug)
+	}
+	return slugs
+}
+
+func codexReasoningEffortsForTest(levels []codexReasoningLevelForTest) []string {
+	efforts := make([]string, 0, len(levels))
+	for _, level := range levels {
+		efforts = append(efforts, level.Effort)
+	}
+	return efforts
+}
+
 func TestGatewayModels_GeminiGroupFallsBackToGeminiModels(t *testing.T) {
 	gin.SetMode(gin.TestMode)
 
@@ -102,9 +371,38 @@ func TestGatewayModels_GeminiGroupFallsBackToGeminiModels(t *testing.T) {
 }
 
 func TestGatewayModels_Grok45AdvertisesReasoningEffortForGrokBuild(t *testing.T) {
+	assertGrokGatewayReasoningEfforts(t, 4409, "grok-4.5", []gatewayReasoningEffortOptionForTest{
+		{Value: "low", Label: "Low"},
+		{Value: "medium", Label: "Medium"},
+		{Value: "high", Label: "High", Default: true},
+	})
+}
+
+func TestGatewayModels_Grok46AdvertisesXHighReasoningEffortForGrokBuild(t *testing.T) {
+	xhighEfforts := []gatewayReasoningEffortOptionForTest{
+		{Value: "low", Label: "Low"},
+		{Value: "medium", Label: "Medium"},
+		{Value: "high", Label: "High", Default: true},
+		{Value: "xhigh", Label: "xHigh"},
+	}
+	tests := []struct {
+		groupID int64
+		model   string
+	}{
+		{groupID: 4410, model: "grok-4.6"},
+		{groupID: 4411, model: "grok-4.6-latest"},
+	}
+	for _, tt := range tests {
+		t.Run(tt.model, func(t *testing.T) {
+			assertGrokGatewayReasoningEfforts(t, tt.groupID, tt.model, xhighEfforts)
+		})
+	}
+}
+
+func assertGrokGatewayReasoningEfforts(t *testing.T, groupID int64, modelID string, want []gatewayReasoningEffortOptionForTest) {
+	t.Helper()
 	gin.SetMode(gin.TestMode)
 
-	groupID := int64(4409)
 	h := newGatewayModelsHandlerForTest(
 		&gatewayModelsAccountRepoStub{
 			byGroup: map[int64][]service.Account{
@@ -113,7 +411,7 @@ func TestGatewayModels_Grok45AdvertisesReasoningEffortForGrokBuild(t *testing.T)
 						ID:       1,
 						Platform: service.PlatformGrok,
 						Credentials: map[string]any{
-							"model_mapping": map[string]any{"grok-4.5": "grok-4.5"},
+							"model_mapping": map[string]any{modelID: modelID},
 						},
 					},
 				},
@@ -135,14 +433,10 @@ func TestGatewayModels_Grok45AdvertisesReasoningEffortForGrokBuild(t *testing.T)
 	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &got))
 	require.Len(t, got.Data, 1)
 	model := got.Data[0]
-	require.Equal(t, "grok-4.5", model.ID)
+	require.Equal(t, modelID, model.ID)
 	require.True(t, model.SupportsReasoningEffort)
 	require.Equal(t, "high", model.ReasoningEffort)
-	require.Equal(t, []gatewayReasoningEffortOptionForTest{
-		{Value: "low", Label: "Low"},
-		{Value: "medium", Label: "Medium"},
-		{Value: "high", Label: "High", Default: true},
-	}, model.ReasoningEfforts)
+	require.Equal(t, want, model.ReasoningEfforts)
 }
 
 func TestGatewayModels_GeminiGroupFiltersMappedModelsByPlatform(t *testing.T) {
@@ -192,6 +486,62 @@ func TestGatewayModels_GeminiGroupFiltersMappedModelsByPlatform(t *testing.T) {
 	require.Equal(t, []string{"gemini-2.5-flash"}, modelIDsForTest(got.Data))
 }
 
+// Scenario: a Composite group with only Anthropic accounts must not inherit Antigravity Gemini defaults.
+func TestGatewayCodexModels_CompositeAnthropicDoesNotAdvertiseAntigravityDefaults(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+
+	groupID := int64(64)
+	h := newGatewayModelsHandlerForTest(&gatewayModelsAccountRepoStub{
+		byGroup: map[int64][]service.Account{
+			groupID: {{ID: 1, Platform: service.PlatformAnthropic}},
+		},
+	})
+
+	rec := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(rec)
+	c.Request = httptest.NewRequest(http.MethodGet, "/models?client_version=0.147.0", nil)
+	c.Set(string(middleware2.ContextKeyAPIKey), &service.APIKey{
+		Group: &service.Group{ID: groupID, Platform: service.PlatformComposite},
+	})
+
+	h.CodexModels(c)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	var got codexModelsResponseForTest
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &got))
+	slugs := codexModelSlugsForTest(got.Models)
+	require.Contains(t, slugs, "claude-opus-4-6")
+	require.NotContains(t, slugs, "gemini-2.5-flash")
+}
+
+// Scenario: Antigravity retains its own Claude and Gemini defaults inside Composite groups.
+func TestGatewayModels_CompositeAntigravityAdvertisesAntigravityDefaults(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+
+	groupID := int64(65)
+	h := newGatewayModelsHandlerForTest(&gatewayModelsAccountRepoStub{
+		byGroup: map[int64][]service.Account{
+			groupID: {{ID: 1, Platform: service.PlatformAntigravity}},
+		},
+	})
+
+	rec := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(rec)
+	c.Request = httptest.NewRequest(http.MethodGet, "/v1/models", nil)
+	c.Set(string(middleware2.ContextKeyAPIKey), &service.APIKey{
+		Group: &service.Group{ID: groupID, Platform: service.PlatformComposite},
+	})
+
+	h.Models(c)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	var got gatewayModelsResponseForTest
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &got))
+	ids := modelIDsForTest(got.Data)
+	require.Contains(t, ids, "claude-opus-4-6")
+	require.Contains(t, ids, "gemini-2.5-flash")
+}
+
 func TestGatewayModels_CustomModelsListDisabledKeepsOriginalModels(t *testing.T) {
 	gin.SetMode(gin.TestMode)
 
@@ -321,6 +671,27 @@ func TestGatewayModels_CompositeCustomModelsListFiltersAcrossConcretePlatforms(t
 							},
 						},
 					},
+					{
+						ID:       4,
+						Platform: service.PlatformKimi,
+						Credentials: map[string]any{
+							"model_mapping": map[string]any{"kimi-custom": "kimi-upstream"},
+						},
+					},
+					{
+						ID:       5,
+						Platform: service.PlatformZhipu,
+						Credentials: map[string]any{
+							"model_mapping": map[string]any{"glm-custom": "glm-upstream"},
+						},
+					},
+					{
+						ID:       6,
+						Platform: service.PlatformDeepseek,
+						Credentials: map[string]any{
+							"model_mapping": map[string]any{"deepseek-custom": "deepseek-upstream"},
+						},
+					},
 				},
 			},
 		},
@@ -335,7 +706,7 @@ func TestGatewayModels_CompositeCustomModelsListFiltersAcrossConcretePlatforms(t
 			Platform: service.PlatformComposite,
 			ModelsListConfig: service.GroupModelsListConfig{
 				Enabled: true,
-				Models:  []string{"gemini-2.5-flash", "missing-model", "ag-custom-model", "gpt-5.5"},
+				Models:  []string{"gemini-2.5-flash", "missing-model", "ag-custom-model", "gpt-5.5", "kimi-custom", "glm-custom", "deepseek-custom"},
 			},
 		},
 	})
@@ -346,7 +717,7 @@ func TestGatewayModels_CompositeCustomModelsListFiltersAcrossConcretePlatforms(t
 
 	var got gatewayModelsResponseForTest
 	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &got))
-	require.Equal(t, []string{"gemini-2.5-flash", "ag-custom-model", "gpt-5.5"}, modelIDsForTest(got.Data))
+	require.Equal(t, []string{"gemini-2.5-flash", "ag-custom-model", "gpt-5.5", "kimi-custom", "glm-custom", "deepseek-custom"}, modelIDsForTest(got.Data))
 }
 
 func TestGatewayModels_CompositeUnmappedAccountsFallbackToLinkedPlatformsOnly(t *testing.T) {
@@ -385,6 +756,139 @@ func TestGatewayModels_CompositeUnmappedAccountsFallbackToLinkedPlatformsOnly(t
 	require.NotContains(t, ids, "gemini-2.5-flash")
 }
 
+// CN 供应商没有静态默认模型列表：composite 下无映射的可调度 CN 账号不得把
+// defaultModelIDsForPlatform default 分支的 Claude 列表挂到 CN 平台名下。
+func TestGatewayModels_CompositeUnmappedCNAccountsContributeNoDefaults(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+
+	groupID := int64(35)
+	h := newGatewayModelsHandlerForTest(
+		&gatewayModelsAccountRepoStub{
+			byGroup: map[int64][]service.Account{
+				groupID: {
+					{ID: 1, Platform: service.PlatformOpenAI},
+					{ID: 2, Platform: service.PlatformKimi},
+					{ID: 3, Platform: service.PlatformZhipu},
+					{ID: 4, Platform: service.PlatformDeepseek},
+				},
+			},
+		},
+	)
+
+	rec := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(rec)
+	c.Request = httptest.NewRequest(http.MethodGet, "/v1/models", nil)
+	c.Set(string(middleware2.ContextKeyAPIKey), &service.APIKey{
+		Group: &service.Group{ID: groupID, Platform: service.PlatformComposite},
+	})
+
+	h.Models(c)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+
+	var got gatewayModelsResponseForTest
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &got))
+
+	ids := modelIDsForTest(got.Data)
+	require.Contains(t, ids, "gpt-5.5")
+	require.NotContains(t, ids, "claude-sonnet-4-6")
+}
+
+// 独立 CN 分组沿用 default 分支的 Claude 默认列表（Claude Code 客户端请求的
+// 就是这些模型名并经账号 model_mapping 转换），composite 支持不得改变该回退。
+func TestDefaultModelIDsForPlatform_CNProvidersKeepClaudeDefaults(t *testing.T) {
+	want := make([]string, 0, len(claude.DefaultModels))
+	for _, model := range claude.DefaultModels {
+		want = append(want, model.ID)
+	}
+	for _, platform := range []string{service.PlatformKimi, service.PlatformZhipu, service.PlatformDeepseek} {
+		require.Equal(t, want, defaultModelIDsForPlatform(platform), "platform=%s", platform)
+	}
+}
+
+func TestDefaultCodexModelIDsForPlatform_DeepSeekUsesDeepSeekModels(t *testing.T) {
+	require.Equal(t, []string{"deepseek-v4-pro", "deepseek-v4-flash"}, defaultCodexModelIDsForPlatform(service.PlatformDeepseek))
+	require.Equal(t, defaultModelIDsForPlatform(service.PlatformAnthropic), defaultCodexModelIDsForPlatform(service.PlatformAnthropic))
+}
+
+func TestGatewayCodexModels_DeepSeekWithoutMappingUsesDeepSeekDefaults(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	const groupID int64 = 130
+	h := newGatewayModelsHandlerForTest(&gatewayModelsAccountRepoStub{
+		byGroup: map[int64][]service.Account{
+			groupID: {
+				{
+					ID:          1,
+					Platform:    service.PlatformDeepseek,
+					Status:      service.StatusActive,
+					Schedulable: true,
+					Credentials: map[string]any{},
+				},
+			},
+		},
+	})
+
+	rec := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(rec)
+	c.Request = httptest.NewRequest(http.MethodGet, "/models?client_version=0.150.0", nil)
+	c.Set(string(middleware2.ContextKeyAPIKey), &service.APIKey{
+		Group: &service.Group{ID: groupID, Platform: service.PlatformDeepseek},
+	})
+
+	h.CodexModels(c)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	var got codexModelsResponseForTest
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &got))
+	slugs := make([]string, 0, len(got.Models))
+	for _, model := range got.Models {
+		slugs = append(slugs, model.Slug)
+	}
+	require.Contains(t, slugs, "deepseek-v4-pro")
+	require.Contains(t, slugs, "deepseek-v4-flash")
+	require.NotContains(t, slugs, "claude-sonnet-4-6")
+	require.NotContains(t, slugs, "claude-opus-4-6")
+}
+
+func TestGatewayCodexModels_OmitsWildcardMappingKeys(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	const groupID int64 = 131
+	h := newGatewayModelsHandlerForTest(&gatewayModelsAccountRepoStub{
+		byGroup: map[int64][]service.Account{
+			groupID: {
+				{
+					ID:       1,
+					Platform: service.PlatformDeepseek,
+					Credentials: map[string]any{
+						"model_mapping": map[string]any{
+							"foo-*":           "deepseek-v4-pro",
+							"deepseek-v4-pro": "deepseek-v4-pro",
+						},
+					},
+				},
+			},
+		},
+	})
+
+	rec := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(rec)
+	c.Request = httptest.NewRequest(http.MethodGet, "/models?client_version=0.150.0", nil)
+	c.Set(string(middleware2.ContextKeyAPIKey), &service.APIKey{
+		Group: &service.Group{ID: groupID, Platform: service.PlatformDeepseek},
+	})
+
+	h.CodexModels(c)
+
+	require.Equal(t, http.StatusOK, rec.Code)
+	var got codexModelsResponseForTest
+	require.NoError(t, json.Unmarshal(rec.Body.Bytes(), &got))
+	slugs := make([]string, 0, len(got.Models))
+	for _, model := range got.Models {
+		slugs = append(slugs, model.Slug)
+	}
+	require.Equal(t, []string{"deepseek-v4-pro"}, slugs)
+}
+
 func TestGatewayModels_CustomModelsListKeepsConcreteModelAllowedByWildcardMapping(t *testing.T) {
 	gin.SetMode(gin.TestMode)
 

backend/internal/handler/gemini_v1beta_handler.go

@@ -569,26 +569,25 @@ func (h *GatewayHandler) GeminiV1BetaModels(c *gin.Context) {
 		forceCacheBilling := fs.ForceCacheBilling
 		quotaPlatform := service.QuotaPlatform(c.Request.Context(), apiKey)
 		sessionID := service.ExtractClientSessionID(c)
+		// 长上下文阶梯由目录数据驱动，统一在计费路径内生效，入口无需声明。
 		h.submitUsageRecordTask(c.Request.Context(), func(ctx context.Context) {
-			if err := h.gatewayService.RecordUsageWithLongContext(ctx, &service.RecordUsageLongContextInput{
-				Result:                result,
-				QuotaPlatform:         quotaPlatform,
-				APIKey:                apiKey,
-				User:                  apiKey.User,
-				Account:               account,
-				Subscription:          subscription,
-				PricingAt:             pricingAt,
-				InboundEndpoint:       inboundEndpoint,
-				UpstreamEndpoint:      upstreamEndpoint,
-				UserAgent:             userAgent,
-				IPAddress:             clientIP,
-				RequestPayloadHash:    requestPayloadHash,
-				LongContextThreshold:  200000, // Gemini 200K 阈值
-				LongContextMultiplier: 2.0,    // 超出部分双倍计费
-				ForceCacheBilling:     forceCacheBilling,
-				APIKeyService:         h.apiKeyService,
-				SessionID:             sessionID,
-				ChannelUsageFields:    clientRequestedUsageFields(c, channelMapping, reqModel, result.UpstreamModel),
+			if err := h.gatewayService.RecordUsage(ctx, &service.RecordUsageInput{
+				Result:             result,
+				QuotaPlatform:      quotaPlatform,
+				APIKey:             apiKey,
+				User:               apiKey.User,
+				Account:            account,
+				Subscription:       subscription,
+				PricingAt:          pricingAt,
+				InboundEndpoint:    inboundEndpoint,
+				UpstreamEndpoint:   upstreamEndpoint,
+				UserAgent:          userAgent,
+				IPAddress:          clientIP,
+				RequestPayloadHash: requestPayloadHash,
+				ForceCacheBilling:  forceCacheBilling,
+				APIKeyService:      h.apiKeyService,
+				SessionID:          sessionID,
+				ChannelUsageFields: clientRequestedUsageFields(c, channelMapping, reqModel, result.UpstreamModel),
 			}); err != nil {
 				logger.L().With(
 					zap.String("component", "handler.gemini_v1beta.models"),

backend/internal/handler/grok_audio.go

@@ -44,39 +44,84 @@ func (h *OpenAIGatewayHandler) GrokRealtime(c *gin.Context) {
 		return
 	}
 
-	selection, _, err := h.gatewayService.SelectAccountWithSchedulerForCapability(
-		c.Request.Context(),
-		apiKey.GroupID,
-		"",
-		"",
-		"grok-4.5",
-		nil,
-		service.OpenAIUpstreamTransportHTTPSSE,
-		// Grok only advertises chat_completions + media capabilities on HEAD.
-		service.OpenAIEndpointCapabilityChatCompletions,
-		false,
-		false,
-		false,
-		service.PlatformGrok,
-	)
-	if err != nil || selection == nil || selection.Account == nil {
-		h.errorResponse(c, http.StatusServiceUnavailable, "api_error", "No available Grok accounts")
-		return
-	}
-
-	var streamStarted bool
 	reqLog := requestLogger(c, "handler.openai_gateway.grok_realtime")
-	release, slotStatus := h.acquireResponsesAccountSlot(c, apiKey.GroupID, "", selection, true, &streamStarted, reqLog)
-	if slotStatus != openAISlotAcquireOK {
+	model := c.Query("model")
+	if strings.TrimSpace(model) == "" {
+		model = "grok-voice-latest"
+	}
+	// Keep the HTTP response uncommitted while selecting and probing an account.
+	// Realtime is not an HTTP streaming response; using reqStream=true here would
+	// let the wait queue flush an SSE ping before the WebSocket handshake succeeds.
+	failed := map[int64]struct{}{}
+	var selection *service.AccountSelectionResult
+	var release func()
+	var token string
+	var upstream *service.GrokRealtimeUpstream
+	var candidateSeen bool
+	for attempts := 0; attempts < 4; attempts++ {
+		// Realtime's voice model is not a text-model capability. Passing a
+		// concrete text model here would reject accounts mapped only to an
+		// older/default text model before the upstream handshake can decide.
+		// An empty requested model keeps account selection capability-based;
+		// the actual voice model remains in the upstream WS query below.
+		candidate, _, selectErr := h.gatewayService.SelectAccountWithSchedulerForCapability(
+			c.Request.Context(), apiKey.GroupID, "", "", "", failed,
+			service.OpenAIUpstreamTransportHTTPSSE,
+			service.OpenAIEndpointCapabilityChatCompletions,
+			false, false, false, service.PlatformGrok,
+		)
+		if selectErr != nil || candidate == nil || candidate.Account == nil {
+			break
+		}
+		candidateSeen = true
+		account := candidate.Account
+		var streamStarted bool
+		var slotStatus openAISlotAcquireResult
+		release, slotStatus = h.acquireResponsesAccountSlot(c, apiKey.GroupID, "", candidate, false, &streamStarted, reqLog)
+		if slotStatus != openAISlotAcquireOK {
+			if slotStatus == openAISlotAcquireFailed {
+				return
+			}
+			failed[account.ID] = struct{}{}
+			continue
+		}
+		var credErr error
+		token, _, credErr = h.gatewayService.GetRequestCredential(c.Request.Context(), c, account)
+		if credErr != nil {
+			release()
+			release = nil
+			failed[account.ID] = struct{}{}
+			continue
+		}
+		probeCtx, cancelProbe := context.WithTimeout(c.Request.Context(), service.DefaultGrokRealtimeDialTimeout)
+		candidateUpstream, openErr := h.gatewayService.OpenGrokRealtime(probeCtx, account, token, model)
+		cancelProbe()
+		if openErr != nil {
+			reqLog.Warn("grok_realtime.pre_accept_failed", zap.Int64("account_id", account.ID), zap.Error(openErr))
+			statusCode := http.StatusBadGateway
+			var dialErr *service.GrokRealtimeDialError
+			if errors.As(openErr, &dialErr) && dialErr.StatusCode > 0 {
+				statusCode = dialErr.StatusCode
+			}
+			h.gatewayService.HandleGrokRealtimeUpstreamError(c.Request.Context(), account, statusCode, []byte(openErr.Error()))
+			release()
+			release = nil
+			failed[account.ID] = struct{}{}
+			continue
+		}
+		selection, upstream = candidate, candidateUpstream
+		break
+	}
+	if selection == nil || selection.Account == nil || release == nil || upstream == nil {
+		if !candidateSeen {
+			h.errorResponse(c, http.StatusServiceUnavailable, "api_error", "No available Grok accounts")
+		} else {
+			h.errorResponse(c, http.StatusBadGateway, "upstream_error", "Grok realtime upstream unavailable")
+		}
 		return
 	}
 	defer release()
-
-	token, _, err := h.gatewayService.GetRequestCredential(c.Request.Context(), c, selection.Account)
-	if err != nil {
-		h.errorResponse(c, http.StatusBadGateway, "upstream_error", "Grok credential unavailable")
-		return
-	}
+	defer func() { _ = upstream.Close() }()
 
 	conn, err := coderws.Accept(c.Writer, c.Request, &coderws.AcceptOptions{CompressionMode: coderws.CompressionContextTakeover})
 	if err != nil {
@@ -84,12 +129,8 @@ func (h *OpenAIGatewayHandler) GrokRealtime(c *gin.Context) {
 	}
 	defer func() { _ = conn.CloseNow() }()
 
-	model := c.Query("model")
-	if strings.TrimSpace(model) == "" {
-		model = "grok-voice-latest"
-	}
 	started := time.Now()
-	audioObserved, proxyErr := h.gatewayService.ProxyGrokRealtime(c.Request.Context(), c, conn, selection.Account, token, model)
+	audioObserved, proxyErr := h.gatewayService.ProxyGrokRealtimeConn(c.Request.Context(), c, conn, upstream)
 	elapsed := time.Since(started)
 	if proxyErr != nil {
 		reqLog.Info("grok_realtime.proxy_failed", zap.Error(proxyErr))

backend/internal/handler/grok_media.go

@@ -346,7 +346,7 @@ func (h *OpenAIGatewayHandler) handleGrokMedia(c *gin.Context, endpoint service.
 					return
 				}
 				if failoverErr.ShouldReportAccountScheduleFailure() {
-					h.gatewayService.ReportOpenAIAccountScheduleResult(account.ID, grokMediaScheduleModel(account, routingModel, nil), false, nil)
+					h.gatewayService.ReportOpenAIAccountScheduleResult(account, grokMediaScheduleModel(account, routingModel, nil), false, nil)
 				}
 				if c.Writer.Size() != writerSizeBeforeForward {
 					h.handleFailoverExhausted(c, failoverErr, true)
@@ -361,19 +361,21 @@ func (h *OpenAIGatewayHandler) handleGrokMedia(c *gin.Context, endpoint service.
 					return
 				}
 				if failoverErr.RetryableOnSameAccount {
-					retryLimit := account.GetPoolModeRetryCount()
-					if sameAccountRetryCount[account.ID] < retryLimit {
+					retryLimit := effectiveSameAccountRetryLimit(failoverErr, account)
+					if sameAccountRetryAllowed(failoverErr, sameAccountRetryCount[account.ID], retryLimit) {
 						sameAccountRetryCount[account.ID]++
+						retryDelay := sameAccountRetryDelayFor(failoverErr, sameAccountRetryCount[account.ID])
 						reqLog.Warn("grok_media.pool_mode_same_account_retry",
 							zap.Int64("account_id", account.ID),
 							zap.Int("upstream_status", failoverErr.StatusCode),
 							zap.Int("retry_limit", retryLimit),
 							zap.Int("retry_count", sameAccountRetryCount[account.ID]),
+							zap.Duration("retry_delay", retryDelay),
 						)
 						select {
 						case <-requestCtx.Done():
 							return
-						case <-time.After(sameAccountRetryDelay):
+						case <-time.After(retryDelay):
 						}
 						continue
 					}
@@ -398,7 +400,7 @@ func (h *OpenAIGatewayHandler) handleGrokMedia(c *gin.Context, endpoint service.
 				)
 				continue
 			}
-			h.gatewayService.ReportOpenAIAccountScheduleResult(account.ID, grokMediaScheduleModel(account, routingModel, nil), false, nil)
+			h.gatewayService.ReportOpenAIAccountScheduleResult(account, grokMediaScheduleModel(account, routingModel, nil), false, nil)
 			if !service.IsResponseCommitted(c) && c.Writer.Size() == writerSizeBeforeForward {
 				h.errorResponse(c, http.StatusBadGateway, "upstream_error", "Upstream request failed")
 			}
@@ -409,7 +411,7 @@ func (h *OpenAIGatewayHandler) handleGrokMedia(c *gin.Context, endpoint service.
 			return
 		}
 
-		h.gatewayService.ReportOpenAIAccountScheduleResult(account.ID, grokMediaScheduleModel(account, routingModel, result), true, nil)
+		h.gatewayService.ReportOpenAIAccountScheduleResult(account, grokMediaScheduleModel(account, routingModel, result), true, nil)
 		if isGrokVideoCreateEndpoint(endpoint) && strings.TrimSpace(result.ResponseID) != "" {
 			if err := h.gatewayService.BindGrokMediaVideoRequestAccount(
 				requestCtx, apiKey.GroupID, result.ResponseID, subject.UserID, apiKey.ID, account.ID,

backend/internal/handler/handler.go

@@ -31,6 +31,7 @@ type AdminHandlers struct {
 	UserAttribute          *admin.UserAttributeHandler
 	ErrorPassthrough       *admin.ErrorPassthroughHandler
 	TLSFingerprintProfile  *admin.TLSFingerprintProfileHandler
+	Plugin                 *admin.PluginHandler
 	APIKey                 *admin.AdminAPIKeyHandler
 	ScheduledTest          *admin.ScheduledTestHandler
 	Channel                *admin.ChannelHandler

backend/internal/handler/model_plaza_handler.go

@@ -15,41 +15,63 @@ import (
 // 广场路由挂 OptionalJWT 中间件：匿名可访问（除非 require_auth 开启），带 token 则
 // 识别用户。可见性规则（橱窗语义，与「可用渠道」的可绑定语义不同）：
 //   - 匿名：仅非专属分组（订阅型照常展示）；
-//   - 登录：非专属分组 + user_allowed_groups 授权的专属分组（不检查订阅有效性）。
+//   - 登录：非专属分组 + user_allowed_groups 授权的专属分组（不检查订阅有效性）；
+//     若该用户开启了公开分组限制，则公开分组同样需要落在授权集合内。
 type ModelPlazaHandler struct {
-	channelService *service.ChannelService
+	plazaService   *service.ModelPlazaService
 	apiKeyService  *service.APIKeyService
 	settingService *service.SettingService
 }
 
 // NewModelPlazaHandler 创建模型广场 handler。
 func NewModelPlazaHandler(
-	channelService *service.ChannelService,
+	plazaService *service.ModelPlazaService,
 	apiKeyService *service.APIKeyService,
 	settingService *service.SettingService,
 ) *ModelPlazaHandler {
 	return &ModelPlazaHandler{
-		channelService: channelService,
+		plazaService:   plazaService,
 		apiKeyService:  apiKeyService,
 		settingService: settingService,
 	}
 }
 
-// modelPlazaOfficialPricing LiteLLM 官方参考价（USD per token）。
+// modelPlazaOfficialPricing 官方参考价（USD per token，与计费目录同源）。
 type modelPlazaOfficialPricing struct {
 	InputPrice        *float64 `json:"input_price"`
 	OutputPrice       *float64 `json:"output_price"`
 	CacheWritePrice   *float64 `json:"cache_write_price"`
 	CacheWrite1hPrice *float64 `json:"cache_write_1h_price,omitempty"`
 	CacheReadPrice    *float64 `json:"cache_read_price"`
+	// Intervals 官方长上下文阶梯，仅多档模型给出。
+	Intervals []userPricingIntervalDTO `json:"intervals,omitempty"`
 }
 
-// modelPlazaModel 广场模型条目：渠道定价（白名单形态）+ 官方参考价。
+// modelPlazaTimePricingPeriod 分时倍率时段（配置时区当天 [start, end)）。
+type modelPlazaTimePricingPeriod struct {
+	StartTime  string  `json:"start_time"`
+	EndTime    string  `json:"end_time"`
+	Multiplier float64 `json:"multiplier"`
+}
+
+// modelPlazaTimePricing 计费会生效的分时倍率（仅倍率 ≠ 1 的时段）。
+// WeekdaysOnly 为 true 时时段仅周一至周五生效，周末整天按标准价计费。
+type modelPlazaTimePricing struct {
+	Timezone     string                        `json:"timezone"`
+	WeekdaysOnly bool                          `json:"weekdays_only,omitempty"`
+	Periods      []modelPlazaTimePricingPeriod `json:"periods"`
+}
+
+// modelPlazaModel 广场模型条目：实收口径展示定价（白名单形态）+ 官方参考价。
 type modelPlazaModel struct {
 	Name            string                     `json:"name"`
 	Platform        string                     `json:"platform"`
 	Pricing         *userSupportedModelPricing `json:"pricing"`
 	OfficialPricing *modelPlazaOfficialPricing `json:"official_pricing"`
+	// LongContextBasis 多档时的计价基准："whole_request"（整单按档）| "marginal"（仅超出部分）。
+	LongContextBasis string `json:"long_context_basis,omitempty"`
+	// TimePricing 分时倍率时段，落在时段内的请求整单乘倍率；无分时省略。
+	TimePricing *modelPlazaTimePricing `json:"time_pricing,omitempty"`
 }
 
 // modelPlazaGroup 广场分组条目（白名单字段）。
@@ -68,9 +90,11 @@ type modelPlazaGroup struct {
 	IsExclusive        bool     `json:"is_exclusive"`
 	// 生图独立倍率：为 true 时图片计费模型的实付倍率取 ImageRateMultiplier，
 	// 不取分组/用户专属倍率。
-	ImageRateIndependent bool              `json:"image_rate_independent"`
-	ImageRateMultiplier  float64           `json:"image_rate_multiplier"`
-	Models               []modelPlazaModel `json:"models"`
+	ImageRateIndependent bool    `json:"image_rate_independent"`
+	ImageRateMultiplier  float64 `json:"image_rate_multiplier"`
+	// 分组是否启用长上下文阶梯计费；关闭时模型实付列只展示最低档/基础价。
+	LongContextPricingEnabled bool              `json:"long_context_pricing_enabled"`
+	Models                    []modelPlazaModel `json:"models"`
 }
 
 // modelPlazaResponse 广场页响应。
@@ -98,17 +122,18 @@ func (h *ModelPlazaHandler) Get(c *gin.Context) {
 		return
 	}
 
-	groups, err := h.channelService.ListPlazaGroups(c.Request.Context())
+	groups, err := h.plazaService.ListGroups(c.Request.Context())
 	if err != nil {
 		response.ErrorFrom(c, err)
 		return
 	}
 
-	// allowedExclusive == nil 表示匿名；登录用户恒为非 nil（可能为空集合）。
-	var allowedExclusive map[int64]struct{}
+	// allowedGroups == nil 表示匿名；登录用户恒为非 nil（可能为空集合）。
+	var allowedGroups map[int64]struct{}
+	var restrictPublicGroups bool
 	var userRates map[int64]float64
 	if authed {
-		allowedExclusive, err = h.apiKeyService.GetUserAllowedGroupIDSet(c.Request.Context(), subject.UserID)
+		allowedGroups, restrictPublicGroups, err = h.apiKeyService.GetUserGroupVisibility(c.Request.Context(), subject.UserID)
 		if err != nil {
 			// 可见性数据拿不到时不能静默降级成匿名视图（会错漏专属分组），直接报错。
 			response.ErrorFrom(c, err)
@@ -122,7 +147,7 @@ func (h *ModelPlazaHandler) Get(c *gin.Context) {
 		}
 	}
 
-	visible := filterPlazaVisibleGroups(groups, allowedExclusive)
+	visible := filterPlazaVisibleGroups(groups, allowedGroups, restrictPublicGroups)
 
 	out := make([]modelPlazaGroup, 0, len(visible))
 	for i := range visible {
@@ -135,18 +160,21 @@ func (h *ModelPlazaHandler) Get(c *gin.Context) {
 }
 
 // filterPlazaVisibleGroups 按登录态裁剪分组可见性。
-// allowedExclusive == nil 表示匿名（仅非专属）；非 nil 表示登录（非专属 + 授权专属）。
+// allowedGroups == nil 表示匿名（仅非专属）；非 nil 表示登录（非专属 + 授权专属）。
+// restrictPublicGroups 为 true 时，公开分组也必须落在 allowedGroups 内，否则用户会
+// 在广场看到自己实际绑定不了的分组。
 func filterPlazaVisibleGroups(
 	groups []service.PlazaGroup,
-	allowedExclusive map[int64]struct{},
+	allowedGroups map[int64]struct{},
+	restrictPublicGroups bool,
 ) []service.PlazaGroup {
 	visible := make([]service.PlazaGroup, 0, len(groups))
 	for _, g := range groups {
-		if g.IsExclusive {
-			if allowedExclusive == nil {
+		if g.IsExclusive || (restrictPublicGroups && allowedGroups != nil) {
+			if allowedGroups == nil {
 				continue
 			}
-			if _, ok := allowedExclusive[g.ID]; !ok {
+			if _, ok := allowedGroups[g.ID]; !ok {
 				continue
 			}
 		}
@@ -161,27 +189,30 @@ func toModelPlazaGroupDTO(g *service.PlazaGroup, userRates map[int64]float64) mo
 	for i := range g.Models {
 		m := &g.Models[i]
 		models = append(models, modelPlazaModel{
-			Name:            m.Name,
-			Platform:        m.Platform,
-			Pricing:         toUserPricing(m.Pricing),
-			OfficialPricing: toModelPlazaOfficialPricing(m.OfficialPricing),
+			Name:             m.Name,
+			Platform:         m.Platform,
+			Pricing:          toUserPricing(m.Pricing),
+			OfficialPricing:  toModelPlazaOfficialPricing(m.OfficialPricing),
+			LongContextBasis: string(m.LongContextBasis),
+			TimePricing:      toModelPlazaTimePricing(m.TimePricing),
 		})
 	}
 	dto := modelPlazaGroup{
-		ID:                   g.ID,
-		Name:                 g.Name,
-		Description:          g.Description,
-		Platform:             g.Platform,
-		SubscriptionType:     g.SubscriptionType,
-		RateMultiplier:       g.RateMultiplier,
-		PeakRateEnabled:      g.PeakRateEnabled,
-		PeakStart:            g.PeakStart,
-		PeakEnd:              g.PeakEnd,
-		PeakRateMultiplier:   g.PeakRateMultiplier,
-		IsExclusive:          g.IsExclusive,
-		ImageRateIndependent: g.ImageRateIndependent,
-		ImageRateMultiplier:  g.ImageRateMultiplier,
-		Models:               models,
+		ID:                        g.ID,
+		Name:                      g.Name,
+		Description:               g.Description,
+		Platform:                  g.Platform,
+		SubscriptionType:          g.SubscriptionType,
+		RateMultiplier:            g.RateMultiplier,
+		PeakRateEnabled:           g.PeakRateEnabled,
+		PeakStart:                 g.PeakStart,
+		PeakEnd:                   g.PeakEnd,
+		PeakRateMultiplier:        g.PeakRateMultiplier,
+		IsExclusive:               g.IsExclusive,
+		ImageRateIndependent:      g.ImageRateIndependent,
+		ImageRateMultiplier:       g.ImageRateMultiplier,
+		LongContextPricingEnabled: g.LongContextPricingEnabled,
+		Models:                    models,
 	}
 	if rate, ok := userRates[g.ID]; ok {
 		dto.UserRateMultiplier = &rate
@@ -189,6 +220,22 @@ func toModelPlazaGroupDTO(g *service.PlazaGroup, userRates map[int64]float64) mo
 	return dto
 }
 
+// toModelPlazaTimePricing 转换分时倍率；nil 透传（JSON 省略）。
+func toModelPlazaTimePricing(p *service.TimePricingSchedule) *modelPlazaTimePricing {
+	if p == nil || len(p.Periods) == 0 {
+		return nil
+	}
+	periods := make([]modelPlazaTimePricingPeriod, 0, len(p.Periods))
+	for _, period := range p.Periods {
+		periods = append(periods, modelPlazaTimePricingPeriod{
+			StartTime:  period.StartTime,
+			EndTime:    period.EndTime,
+			Multiplier: period.Multiplier,
+		})
+	}
+	return &modelPlazaTimePricing{Timezone: p.Timezone, WeekdaysOnly: p.WeekdaysOnly, Periods: periods}
+}
+
 // toModelPlazaOfficialPricing 转换官方参考价；nil 透传（前端显示 "-"）。
 func toModelPlazaOfficialPricing(p *service.PlazaOfficialPricing) *modelPlazaOfficialPricing {
 	if p == nil {
@@ -200,5 +247,6 @@ func toModelPlazaOfficialPricing(p *service.PlazaOfficialPricing) *modelPlazaOff
 		CacheWritePrice:   p.CacheWritePrice,
 		CacheWrite1hPrice: p.CacheWrite1hPrice,
 		CacheReadPrice:    p.CacheReadPrice,
+		Intervals:         toUserPricingIntervals(p.Intervals),
 	}
 }

backend/internal/handler/model_plaza_handler_test.go

@@ -25,7 +25,7 @@ func plazaGroups() []service.PlazaGroup {
 
 func TestFilterPlazaVisibleGroups_AnonymousSeesOnlyNonExclusive(t *testing.T) {
 	// 匿名(allowedExclusive == nil):仅非专属分组;订阅型公开分组照常可见(橱窗语义)。
-	visible := filterPlazaVisibleGroups(plazaGroups(), nil)
+	visible := filterPlazaVisibleGroups(plazaGroups(), nil, false)
 	require.Len(t, visible, 2)
 	ids := []int64{visible[0].ID, visible[1].ID}
 	require.ElementsMatch(t, []int64{1, 3}, ids)
@@ -34,7 +34,7 @@ func TestFilterPlazaVisibleGroups_AnonymousSeesOnlyNonExclusive(t *testing.T) {
 func TestFilterPlazaVisibleGroups_AuthedSeesGrantedExclusive(t *testing.T) {
 	// 登录:非专属 + 授权的专属;未授权的专属仍不可见。
 	allowed := map[int64]struct{}{2: {}}
-	visible := filterPlazaVisibleGroups(plazaGroups(), allowed)
+	visible := filterPlazaVisibleGroups(plazaGroups(), allowed, false)
 	require.Len(t, visible, 3)
 	ids := make([]int64, 0, len(visible))
 	for _, g := range visible {
@@ -46,10 +46,33 @@ func TestFilterPlazaVisibleGroups_AuthedSeesGrantedExclusive(t *testing.T) {
 func TestFilterPlazaVisibleGroups_AuthedEmptySetSeesNoExclusive(t *testing.T) {
 	// 登录但无任何专属授权(空集合,非 nil):与匿名同样只见非专属,
 	// 但语义区分要保持——空集合不能被当作 nil 匿名分支。
-	visible := filterPlazaVisibleGroups(plazaGroups(), map[int64]struct{}{})
+	visible := filterPlazaVisibleGroups(plazaGroups(), map[int64]struct{}{}, false)
 	require.Len(t, visible, 2)
 }
 
+func TestFilterPlazaVisibleGroups_RestrictedUserSeesOnlyGrantedPublic(t *testing.T) {
+	// 开启公开分组限制后，公开分组也必须落在授权集合内，否则用户会在广场
+	// 看到自己实际绑定不了的分组。
+	allowed := map[int64]struct{}{1: {}, 2: {}}
+	visible := filterPlazaVisibleGroups(plazaGroups(), allowed, true)
+	ids := make([]int64, 0, len(visible))
+	for _, g := range visible {
+		ids = append(ids, g.ID)
+	}
+	// 3 是未授权的公开分组，受限后不可见；4 是未授权的专属分组，一贯不可见。
+	require.ElementsMatch(t, []int64{1, 2}, ids)
+}
+
+func TestFilterPlazaVisibleGroups_RestrictionDoesNotAffectAnonymous(t *testing.T) {
+	// 匿名没有用户记录，限制标志无从谈起，可见性必须与未受限时一致。
+	visible := filterPlazaVisibleGroups(plazaGroups(), nil, true)
+	ids := make([]int64, 0, len(visible))
+	for _, g := range visible {
+		ids = append(ids, g.ID)
+	}
+	require.ElementsMatch(t, []int64{1, 3}, ids)
+}
+
 func TestModelPlazaHandler_NilSettingServiceFailsClosed404(t *testing.T) {
 	gin.SetMode(gin.TestMode)
 	h := &ModelPlazaHandler{} // settingService == nil → fail-closed
@@ -91,7 +114,7 @@ func TestToModelPlazaGroupDTO_UserRateAndFieldWhitelist(t *testing.T) {
 		"id", "name", "description", "platform", "subscription_type",
 		"rate_multiplier", "user_rate_multiplier", "is_exclusive", "models",
 		"peak_rate_enabled", "peak_start", "peak_end", "peak_rate_multiplier",
-		"image_rate_independent", "image_rate_multiplier",
+		"image_rate_independent", "image_rate_multiplier", "long_context_pricing_enabled",
 	} {
 		_, exists := decoded[key]
 		require.Truef(t, exists, "plaza group DTO must expose %q", key)
@@ -109,6 +132,12 @@ func TestToModelPlazaGroupDTO_UserRateAndFieldWhitelist(t *testing.T) {
 	require.Contains(t, official, "cache_read_price")
 	_, has1h := official["cache_write_1h_price"]
 	require.False(t, has1h, "1h 缓存写价为 nil 时应 omitempty")
+	_, hasOfficialIntervals := official["intervals"]
+	require.False(t, hasOfficialIntervals, "官方无阶梯时 intervals 应 omitempty")
+	_, hasBasis := model["long_context_basis"]
+	require.False(t, hasBasis, "单档模型不输出 long_context_basis")
+	_, hasTimePricing := model["time_pricing"]
+	require.False(t, hasTimePricing, "无分时时不输出 time_pricing")
 
 	// 无专属倍率:user_rate_multiplier 整个字段省略
 	dtoNoRate := toModelPlazaGroupDTO(&g, nil)
@@ -124,4 +153,94 @@ func TestToModelPlazaOfficialPricing_NilPassthrough(t *testing.T) {
 	require.Nil(t, toModelPlazaOfficialPricing(nil))
 }
 
+func TestToModelPlazaGroupDTO_LongContextTiersAndBasis(t *testing.T) {
+	maxTokens := 272000
+	g := service.PlazaGroup{
+		ID: 3, Name: "ladder", Platform: "openai", SubscriptionType: "standard", RateMultiplier: 1,
+		LongContextPricingEnabled: true,
+		Models: []service.PlazaModel{{
+			Name:     "gpt-5.4",
+			Platform: "openai",
+			Pricing: &service.ChannelModelPricing{
+				BillingMode: service.BillingModeToken,
+				InputPrice:  testPtr(2.5e-6),
+				Intervals: []service.PricingInterval{
+					{MinTokens: 0, MaxTokens: &maxTokens, TierLabel: "≤272K", InputPrice: testPtr(2.5e-6)},
+					{MinTokens: 272000, TierLabel: ">272K", InputPrice: testPtr(5e-6)},
+				},
+			},
+			OfficialPricing: &service.PlazaOfficialPricing{
+				InputPrice: testPtr(2.5e-6),
+				Intervals: []service.PricingInterval{
+					{MinTokens: 0, MaxTokens: &maxTokens, TierLabel: "≤272K", InputPrice: testPtr(2.5e-6)},
+					{MinTokens: 272000, TierLabel: ">272K", InputPrice: testPtr(5e-6)},
+				},
+			},
+			LongContextBasis: service.ContextPricingBasisWholeRequest,
+		}},
+	}
+
+	raw, err := json.Marshal(toModelPlazaGroupDTO(&g, nil))
+	require.NoError(t, err)
+	var decoded map[string]any
+	require.NoError(t, json.Unmarshal(raw, &decoded))
+	require.Equal(t, true, decoded["long_context_pricing_enabled"])
+
+	model := decoded["models"].([]any)[0].(map[string]any)
+	require.Equal(t, "whole_request", model["long_context_basis"])
+
+	pricing := model["pricing"].(map[string]any)
+	paidTiers := pricing["intervals"].([]any)
+	require.Len(t, paidTiers, 2)
+	require.Equal(t, ">272K", paidTiers[1].(map[string]any)["tier_label"])
+
+	official := model["official_pricing"].(map[string]any)
+	officialTiers := official["intervals"].([]any)
+	require.Len(t, officialTiers, 2)
+	first := officialTiers[0].(map[string]any)
+	require.Equal(t, "≤272K", first["tier_label"])
+	require.InDelta(t, 272000, first["max_tokens"].(float64), 0)
+	require.Contains(t, first, "cache_write_price", "区间 DTO 字段齐全（nil 输出 null）")
+}
+
 func testPtr(v float64) *float64 { return &v }
+
+func TestToModelPlazaGroupDTO_TimePricing(t *testing.T) {
+	g := service.PlazaGroup{
+		ID: 4, Name: "cn", Platform: "deepseek", SubscriptionType: "standard", RateMultiplier: 1,
+		Models: []service.PlazaModel{{
+			Name:     "deepseek-chat",
+			Platform: "deepseek",
+			Pricing:  &service.ChannelModelPricing{BillingMode: service.BillingModeToken, InputPrice: testPtr(0.28e-6)},
+			TimePricing: &service.TimePricingSchedule{Timezone: "Asia/Shanghai", Periods: []service.TimePricingPeriod{
+				{StartTime: "00:30", EndTime: "08:30", Multiplier: 0.5},
+			}},
+		}, {
+			Name:     "deepseek-reasoner",
+			Platform: "deepseek",
+			Pricing:  &service.ChannelModelPricing{BillingMode: service.BillingModeToken, InputPrice: testPtr(0.56e-6)},
+			TimePricing: &service.TimePricingSchedule{Timezone: "Asia/Shanghai", WeekdaysOnly: true, Periods: []service.TimePricingPeriod{
+				{StartTime: "00:30", EndTime: "08:30", Multiplier: 0.5},
+			}},
+		}},
+	}
+	raw, err := json.Marshal(toModelPlazaGroupDTO(&g, nil))
+	require.NoError(t, err)
+	var decoded map[string]any
+	require.NoError(t, json.Unmarshal(raw, &decoded))
+	model := decoded["models"].([]any)[0].(map[string]any)
+	tp := model["time_pricing"].(map[string]any)
+	require.Equal(t, "Asia/Shanghai", tp["timezone"])
+	_, hasWeekdaysOnly := tp["weekdays_only"]
+	require.False(t, hasWeekdaysOnly, "未开启仅工作日时字段省略")
+	periods := tp["periods"].([]any)
+	require.Len(t, periods, 1)
+	first := periods[0].(map[string]any)
+	require.Equal(t, "00:30", first["start_time"])
+	require.Equal(t, "08:30", first["end_time"])
+	require.InDelta(t, 0.5, first["multiplier"].(float64), 1e-12)
+
+	weekdaysModel := decoded["models"].([]any)[1].(map[string]any)
+	weekdaysTP := weekdaysModel["time_pricing"].(map[string]any)
+	require.Equal(t, true, weekdaysTP["weekdays_only"])
+}

backend/internal/handler/no_account_error.go

@@ -4,6 +4,8 @@ import (
 	"context"
 	"fmt"
 	"net/http"
+	"regexp"
+	"strconv"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -32,6 +34,29 @@ type noAccountErrorClassification struct {
 	ModelNotFound bool // true when this is a 404 model_not_found classification
 }
 
+var selectionModelRateLimitedPattern = regexp.MustCompile(`(?:model_rate_limited|rate_limited)=(\d+)`)
+
+// classifySelectionFailureError preserves the scheduler's compact reason when
+// every model-capable account is temporarily rate limited.
+func classifySelectionFailureError(err error, fallback noAccountErrorClassification) noAccountErrorClassification {
+	if err == nil {
+		return fallback
+	}
+	match := selectionModelRateLimitedPattern.FindStringSubmatch(strings.ToLower(err.Error()))
+	if len(match) != 2 {
+		return fallback
+	}
+	count, parseErr := strconv.Atoi(match[1])
+	if parseErr != nil || count <= 0 {
+		return fallback
+	}
+	return noAccountErrorClassification{
+		Status:  http.StatusTooManyRequests,
+		ErrType: "rate_limit_error",
+		Message: "All available accounts are currently rate-limited. Please retry later.",
+	}
+}
+
 // classifyNoAccountError decides between 404 model_not_found and 503
 // api_error for "no available accounts" failures.
 //
@@ -106,7 +131,11 @@ func classifyNoAccountErrorFromGin(
 	if c != nil && c.Request != nil {
 		ctx = c.Request.Context()
 	}
-	return classifyNoAccountError(ctx, diag, apiKey, routingModel, displayModel, platform)
+	classification := classifyNoAccountError(ctx, diag, apiKey, routingModel, displayModel, platform)
+	if classification.ModelNotFound {
+		service.MarkOpsClientBusinessLimited(c, service.OpsClientBusinessLimitedReasonLocalModelConfiguration)
+	}
+	return classification
 }
 
 func classifyOpenAICompatibleNoAccountErrorFromGin(