原型 · 预览版:收录官方注册中心里的全部服务器,每 30 分钟同步一次。评价和账户功能将在后续阶段推出。
PopNx MCP为使用 AI 的人准备的 MCP 服务器 · 用大白话讲清楚

AI Agents API Library

作者:eater2版本 0.3.12i★ –i电脑或云端
注册名称io.github.eater2/ai-agents-api-library

安装

概览

作者提供的说明(英文原文)

Find APIs and MCP servers for AI agents: auth, free tier, MCP endpoint, docs, last-checked date.

PopNx MCP 的三大支柱

这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。

它能帮你做什么

我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。

安装

选择你使用的 AI 应用,然后复制设置。

安装之前

你需要

  • 在云端运行无需安装任何东西,通过互联网连接
  • npm已安装 Node.js(用于 npx 命令)

你用的是哪个应用?

在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。

适用于网页版、桌面版和手机版 Claude:

  1. 打开 Settings → Connectors
  2. 点击 Add custom connector
  3. 粘贴下面的地址,然后点击 Add
  4. 如果要求登录,请登录作者的服务
SettingsConnectors1+ Add2https://…3
https://ai-agents-api-library.vercel.app/mcp

或者在你的电脑上运行

打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json

{
  "mcpServers": {
    "ai-agents-api-library": {
      "command": "npx",
      "args": [
        "-y",
        "ai-agents-api-library"
      ]
    }
  }
}

在终端中运行这条命令

claude mcp add --transport http ai-agents-api-library https://ai-agents-api-library.vercel.app/mcp

或者在你的电脑上运行

claude mcp add ai-agents-api-library -- npx -y ai-agents-api-library

在 Cursor 中打开

或者自己复制设置

粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json

{
  "mcpServers": {
    "ai-agents-api-library": {
      "url": "https://ai-agents-api-library.vercel.app/mcp"
    }
  }
}

或者在你的电脑上运行

在 Cursor 中打开

或者自己复制设置

粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json

{
  "mcpServers": {
    "ai-agents-api-library": {
      "command": "npx",
      "args": [
        "-y",
        "ai-agents-api-library"
      ]
    }
  }
}

浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。

在 VS Code 中打开

或者自己复制设置

粘贴到项目中的 .vscode/mcp.json

{
  "servers": {
    "ai-agents-api-library": {
      "type": "http",
      "url": "https://ai-agents-api-library.vercel.app/mcp"
    }
  }
}

或者在你的电脑上运行

在 VS Code 中打开

或者自己复制设置

粘贴到项目中的 .vscode/mcp.json

{
  "servers": {
    "ai-agents-api-library": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "ai-agents-api-library"
      ]
    }
  }
}

浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。

ChatGPT 只能连接互联网上的服务器:

  1. 打开网页版 ChatGPT → Settings → Connectors (Apps)
  2. 点击 Create 或 Add(部分套餐需要先在 Advanced 中开启 Developer mode)
  3. 粘贴下面的地址,起个名字,然后保存
  4. 如果要求登录,请登录作者的服务
SettingsConnectors1+ Create2https://…3
https://ai-agents-api-library.vercel.app/mcp

菜单名称可能因套餐和版本不同而略有差异。

你的选择通过 Cookie 保存在这台设备上,无需账户。

这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。

它能做什么

工具名称和描述与服务器在 2026年11月6日 报告的完全一致。这些内容由作者撰写,不是我们写的;我们以纯文本形式显示。

  • list_categories

    List all categories in the catalog with the number of services in each.

  • search_apis

    Search 300+ catalog entries of third-party APIs and MCP servers (each with a last-checked date) by operation or keywords, with optional category and quality filters. `match`: exact = the entry lists the operation and its example call or an MCP tool performs it; listed = the entry

  • get_api

    Full entry for one service: docs, auth scheme and how the credential is sent, base URL, operations, MCP endpoint or repository (a repository must be installed; a docs-only server does not call the API), free plan vs trial, rate limits, data policy, notes, last-checked dates, link

  • get_reviews

    Public user reviews of one service (up to 100, from SourceForge): rating, title, pros, cons, overall, reviewer role, company size, date and link to the original. Reviewer names are not included. Reviews do not show whether the API fits the task. Each review has aspect tags; `topi

基础检查i

我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月6日

  • ✓ 通过

    作者与仓库一致

    注册名称已通过 GitHub 账户 eater2 验证,该账户拥有此仓库

  • … 尚未检查

    仍在维护

    我们还没有从 GitHub 读取这个仓库

  • … 尚未检查

    有许可证

    我们还没有从 GitHub 读取这个仓库

  • ✓ 通过

    端点有响应

    在 182 毫秒内响应(HTTP 200)

  • – 不适用

    使用 OAuth 登录

    无需登录

  • ✓ 通过

    工具描述中没有隐藏指令

    已读取 4 个工具描述,未发现针对 AI 的隐藏指令

与“其他”中的其他服务器比较i

  • 星标数多于此分类中其他 32531 个服务器的 0%
  • 发布到注册中心的时间比此分类中其他 32531 个服务器的 93% 更新
  • 基础检查结果等于或优于此分类中其他 32531 个服务器的 100%

与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。

常见问题

AI Agents API Library 安全吗?

它在 3 项适用的基础检查中通过了 3 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。

它需要哪些密钥或代码?

注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。

它在我的电脑上运行还是在云端?

两种方式都可以:安装在你的电脑上,或者连接云端版本。

它还在维护吗?

GitHub 仓库最后更新于–,最新版本(0.3.12)于前天进入注册中心。

哪些应用可以使用它?

支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,所以接受远程 MCP 服务器的应用也可以使用,例如开启开发者模式的 ChatGPT。

有什么可以替代它?

同一分类中有:Spec Driven Development, Remote Desktop Commander, Reelsaw

替代选择

同一分类中其他作者的服务器,按星标数排序。

如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i

Spec Driven Development

作者:formulahendry

MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.

在你的电脑上运行★ 440MIT14个月前更新
安装

Remote Desktop Commander

作者:desktop-commander

Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.

在云端运行★ 211昨天更新
安装

Reelsaw

作者:osbytes

Drive a self-hosted Reelsaw: turn long videos into vertical clips, edit, re-render and export them.

在你的电脑上运行今天更新
安装

作者的 README

打开 README

按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。

Polski · HTML · llms.txt · catalog.json

AI Agents API Library

A curated, machine-readable encyclopedia of third-party APIs and MCP servers that AI agents can call on a user's behalf.

This article is about services an AI agent can operate programmatically, not AI products used by humans through a web UI.

AI Agents API Library Type API & MCP catalog Entries 340 Categories 30 Official MCP servers 238 Languages English, Polish Format JSON, Markdown, HTML, llms.txt Last checked 2026-11-04 License CC BY 4.0 (data), MIT (code)

AI Agents API Library is an open, bilingual (English/Polish) catalog of 340 services (each with a last-checked date) in 30 categories that an autonomous AI agent — such as Claude, ChatGPT, Gemini or a custom LLM agent — can use without a human in the loop, after a one-time human step: registering an account and handing the agent an API key, an OAuth token, or a connection to an MCP (Model Context Protocol) server. Every entry records the documentation URL, the authentication method, the availability of an official or community MCP server, and the free tier. The list is not a directory of products with “AI” in their name; the criterion is agent operability, not branding.

Contents

  1. Machine-readable access (for AI agents)
  2. Inclusion criteria
  3. Legend
  4. Web search (14) · Web scraping and crawling (12) · Browser automation and cloud browsers (7) · Knowledge and research data (19) · Image generation and editing (10) · Video generation and editing (14) · Speech and audio (13) · Music generation (6) · 3D generation and assets (6) · Architecture, CAD and BIM (8) · Diagrams and software architecture (8) · Design and UI (6) · Documents, OCR and presentations (14) · Translation and language (6) · Code execution sandboxes (11) · Developer platforms and DevOps (13) · Cloud and hosting (11) · Databases (12) · Vector search, memory and RAG (9) · Email, team chat and notifications (10) · SMS and messaging apps (12) · Voice agents and telephony (9) · Productivity and workspace (17) · CRM, support and marketing (8) · Social media (10) · Maps, geolocation and weather (18) · Finance, payments and market data (16) · E-commerce (8) · Automation and integration platforms (15) · Model APIs and inference (18)
  5. Related directories and registries
  6. Contributing
  7. See also
  8. References

Machine-readable access (for AI agents)

AI agents should not parse this page. Fetch one of these stable files instead:

FileRaw URL
llms.txt — short index for LLMs, links to everything belowhttps://raw.githubusercontent.com/eater2/ai_agents_api_library/main/llms.txt
llms-full.txt — the whole catalog as compact plain text, one line per servicehttps://raw.githubusercontent.com/eater2/ai_agents_api_library/main/llms-full.txt
catalog/all.json — full catalog, one JSON arrayhttps://raw.githubusercontent.com/eater2/ai_agents_api_library/main/catalog/all.json
catalog/ .json — one file per category, for small context windowshttps://raw.githubusercontent.com/eater2/ai_agents_api_library/main/catalog/
data/schema.json — JSON Schema of an entryhttps://raw.githubusercontent.com/eater2/ai_agents_api_library/main/data/schema.json

Recommended agent workflow: (1) pick the category file, (2) filter by auth and mcp.type, (3) read docs (or llms_txt / openapi when present), (4) ask the user for the credential named in auth_hint.

MCP server

The catalog is also an MCP server with four tools: search_apis (filters: query, category, mcp (official/remote/any), no_auth, free_tier, no_card, auth, min_rating, min_reviews; sort by fit, rating or review count; results carry a reviews_summary of praised and criticised aspects), get_api, get_reviews (user reviews with pros and cons) and list_categories. No key needed. Add it to your MCP client:

{

"mcpServers": { "ai-agents-api-library": { "command": "npx", "args": ["-y", "ai-agents-api-library"] } } }

Claude Code: claude mcp add ai-agents-api-library -- npx -y ai-agents-api-library

No install: remote endpoint (Streamable HTTP, no key) https://ai-agents-api-library.vercel.app/mcp, e.g. claude mcp add --transport http ai-agents-api-library https://ai-agents-api-library.vercel.app/mcp. Listed in the official MCP Registry as io.github.eater2/ai-agents-api-library.

Claude skill and plugin

The skill tells Claude when to reach for the catalog and how to pick a service. The plugin bundles the skill with the MCP server:

  • Claude Code plugin (skill + MCP server):
  /plugin marketplace add eater2/ai_agents_api_library

/plugin install ai-agents-api-library@eater2

- **Skill only, Claude Code:** copy [`plugin/skills/ai-agents-api-library/`](plugin/skills/ai-agents-api-library/SKILL.md) to `~/.claude/skills/` (all projects) or `.claude/skills/` (one project).
  • Skill only, claude.ai / Claude Desktop: download the skill zip and upload it under Settings → Capabilities → Skills.

Inclusion criteria

  • The service exposes a public, self-serve programmatic interface (REST, GraphQL, SDK) or an MCP server.
  • A human needs to act at most once — sign up and create a key or approve OAuth. After that the agent works unattended.
  • The interface is documented and currently active (links last checked on the date shown in the infobox).
  • Consumer-only apps, waitlists, sales-gated enterprise products, agent frameworks and IDEs are excluded — they are agents or tools for building agents, not services an agent calls.

Legend

  • MCP ✅ — official MCP server maintained by the vendor
  • MCP ◐ — community-maintained MCP server
  • MCP — — no MCP server known; use the REST API
  • 🆓 — free tier or free usage without payment
  • ★ (date) — proof under the service name: SourceForge rating and review count, GitHub stars, GitHub issues fixed and days to fix, Smithery uses — linked, with the date fetched

*Ratings and usage (★) fetched: 2026-11-04 09:20 UTC.*

Web search

*Search engines and answer APIs that return fresh web results, snippets or full page content in agent-friendly formats.* — 14 services. JSON: catalog/web-search.json

ServiceWhat an agent can doAuthMCPFree tierDocs

| Brave Search API ⭐⭐⭐⭐⭐ 4.5/5 · 2 reviews [](https://github.com/brave/

只显示了开头部分;其余内容请在仓库中阅读

问 PopNx