AI Rank
← 开源
router-for-me

Cli-Proxy-API-Management-Center

This is a WebUI interface based on CLI-Proxy-API, designed to simplify configuration modifications and runtime status monitoring.

InfrastructureObservabilityGatewayTypeScript
在 GitHub 打开
增长势头
+1924 小时新增 Star+0.4%
4.47k
Star
1.36k
Fork
—
本周
40
贡献者
创建于 2025-09-06 · 更新于 2026-10-05 · 今日第 538 名
主要开发者
LTbinglingfenghkfiresmoxi000
README

CLI Proxy API

CLI Proxy API Management Center

Your proxy, at a glance. Your configuration, in control.

A lightweight Web UI for CLI Proxy API.

Manage providers, credentials, quotas, and logs — from a single HTML file.

Backend: v8 Single-file build License: MIT

English · 简体中文

Get started · Features · Development · Releases

One place to manage your instance

From everyday configuration to troubleshooting, keep the essentials close at hand.

Configure Connect Observe
Edit settings visually or in YAML, with a diff before saving. Manage provider keys, auth files, and OAuth flows. Check instance status, provider quotas, and request logs.

Single-file deployment · Responsive layout · Four UI languages

[!IMPORTANT] Requires CLI Proxy API ≥ 8.0.0; the latest v8 release is recommended. This repository is the management UI, not the proxy — it does not forward traffic. It uses the v8 Management API (/v8/management) and v8 configuration layout, with no v0 fallback. Plugin resources and custom extensions retain their backend-declared paths.

Quick start

The UI is bundled with CLI Proxy API. No separate frontend deployment is needed for the standard setup.

  1. Start your CLI Proxy API service.
  2. Open http://:/management.html.
  3. Enter your management key and connect.

The API address is detected from the page URL and can be changed manually.

[!NOTE] The management key signs you into this UI. Client keys in access.api-keys authorize requests to the proxy — they are not interchangeable.

Connection settings & remote access

The UI accepts addresses such as localhost:8317, https://example.com:8317, or http://example.com:8317/v8/management; the management suffix is removed automatically.

Management requests use Authorization: Bearer . Remote access may require management.allow-remote: true on the server. See the backend documentation for authentication rules and server-side limits.

Upgrading from v7

Upgrade the backend first and back up config.yaml. The backend returns the v8 configuration layout on reads; a successful v8 configuration write migrates the stored file. The editor only writes v8 fields:

  • access.api-keys: client keys for accessing the proxy.
  • Top-level api-keys: upstream provider groups.

Features

Area What you can do
Dashboard See connection status, server version, build date, and model availability at a glance.
Configuration Edit common settings and client keys visually, or use the YAML editor with search, highlighting, and a save diff preview.
AI providers Configure Gemini, Codex, Claude, Vertex, and OpenAI-compatible providers; manage keys, headers, proxies, and model mappings.
Auth files & OAuth Upload, download, and organize credentials; connect supported providers with OAuth or device flows; manage model aliases and exclusions.
Quotas Inspect quota and usage information for supported providers, including Claude, Antigravity, Codex, Kimi, and xAI/Grok.
Logs Follow logs with auto-refresh, search, hide management traffic, and download request error logs.
Plugins Access plugin management when the connected backend advertises support.
System Check for updates, inspect available models, and clear local login data.

Supports English, 简体中文, 繁體中文, and Русский, with browser-language detection and a manual language switch. Responsive layouts support desktop, tablet, and mobile use in modern Chrome, Firefox, Safari, and Edge.

Sponsor

APIMart — AI image and video generation API

Thanks to APIMart for sponsoring this project!

APIMart is a low-cost API platform for AI image & video generation — GPT-Image-2 from $0.006/image, 160+ images per dollar. One async API covers both image and video: submit a task, get an ID, fetch results via polling or callback. Batch tens of thousands of images without timeouts, switch models without changing code. Pay-as-you-go with no monthly fee — sign up here to get started.

Deployment

To build your own single-file UI, use Bun 1.3.14:

bun install --frozen-lockfile
bun run build

The output is dist/index.html, with JavaScript, CSS, and bundled assets inlined. The release workflow renames it to management.html for backend hosting.

Use bun run preview to preview locally. Prefer an HTTP server over opening the file via file://, which can encounter browser CORS restrictions.

Release details

  • Tags matching vX.Y.Z trigger the release workflow.
  • The UI version is injected at build time from VERSION, then git tags, then the package version, with dev as the final fallback.
  • Hash routing and an ES2020 build target keep deployment simple.

Development

bun install --frozen-lockfile
bun run dev

Open http://localhost:5173 and connect to your backend instance.

Built with React 19 · TypeScript 6 · Vite 8 · Zustand · Axios · React Router 7 · Motion · CodeMirror 6 · SCSS Modules · i18next.

Command Purpose
bun run dev Start the Vite dev server
bun run build TypeScript compilation + production build
bun run preview Serve the build locally
bun run test Run Bun tests
bun run lint Run ESLint; warnings are not treated as failures
bun run type-check Run tsc --noEmit
bun run verify Run tests, lint, and build
bun run format Format source files with Prettier

Issues and PRs are welcome. Include reproduction steps, backend and UI versions, screenshots for UI changes, and verification results. See AGENTS.md for repository conventions.

Security

Treat the management key as a secret. When remember password is enabled, it is persisted in browser storage using reversible obfuscation — not encryption.

Use a trusted device or dedicated browser profile. Enable remote management only after evaluating the exposure of your server.

Troubleshooting

Connection or authentication problems

  • Cannot connect / 401: check the API address and management key. Remote connections may require remote management to be enabled.
  • Repeated authentication failures: the server may temporarily block remote IPs.

Missing pages, unsupported features, or failed tests

  • Logs page missing: enable “Logging to file” in Basic Settings.
  • Feature unsupported: check the backend version and whether the relevant endpoint is enabled. Some capabilities depend on backend support.
  • Model list unavailable: querying /v1/models requires at least one proxy API key.
  • OpenAI provider test fails: this test runs in the browser and depends on the provider's network reachability and CORS policy. Failure does not necessarily mean the backend cannot reach it.

CLI Proxy API · Report an issue · MIT License