NetReach
Provides AI agents with read-only access to various public data sources (web, YouTube, RSS, GitHub, V2EX, Bilibili, and semantic search) without requiring any login credentials or API keys.
README
👁️ NetReach — no-login 版
给你的 AI Agent 一键装上互联网「只读 + 搜索」能力 —— 只收录无需登录、爬虫风险低的公开数据源。
支持的平台(全部零配置,无需登录)
| 平台 | 能力 | 后端 |
|---|---|---|
| 🌐 网页 | 读任意网页 → Markdown | Jina Reader(curl r.jina.ai/URL) |
| 📺 YouTube | 字幕提取 + 视频搜索 | yt-dlp |
| 📡 RSS | 解析任意 RSS/Atom 源 | feedparser |
| 📦 GitHub | 公开仓库 + 代码搜索 | gh CLI(公开数据无需登录) |
| 💻 V2EX | 热门/节点/帖子详情+回复/用户 | 公开 API |
| 📺 B站 | 搜索 + 视频详情(无需登录) | bili-cli ▸ 搜索 API |
| 🔍 全网搜索 | AI 语义搜索 | Exa(mcporter MCP,免费无需 Key) |
⚠️ 注意:B站字幕需要浏览器登录态,不在本版本范围内(不登录只能搜视频、看详情)。
设计理念
NetReach 是一个能力层(capability layer),不是又一个工具。
它负责选型、安装、体检、路由,不负责底层读取本身。读取由 Agent 直接调用上游工具(yt-dlp、gh、feedparser、bili-cli 等)完成,没有包装层。
每个平台 = 首选 + 备选的有序后端列表。换接入方式 = 调整列表顺序,不是重写代码。agent-reach doctor 会告诉你每个平台当前在用哪个后端。
channels/
├── web.py → Jina Reader
├── youtube.py → yt-dlp
├── rss.py → feedparser
├── github.py → gh CLI
├── v2ex.py → V2EX 公开 API(直接 HTTP 实现)
├── bilibili.py → bili-cli ▸ B站搜索 API
├── exa_search.py → Exa via mcporter
└── __init__.py → 渠道注册
每个渠道文件按序真实探测各候选后端(不只是看命令存不存在),第一个完整可用的当选;坏掉的会给出修复处方。
快速上手
# 安装(默认安全检查,只读不改系统)
agent-reach install --env=auto
# 显式授权安装系统依赖 + 注册 skill
agent-reach install --env=auto --system
# 体检:一条命令看每个渠道通不通、走哪个后端
agent-reach doctor
安装完直接告诉 Agent 就行,不用记命令:
- "看看这个链接" →
curl https://r.jina.ai/URL - "这个 GitHub 仓库是干嘛的" →
gh repo view owner/repo - "这个 YouTube 视频讲了什么" →
yt-dlp提取字幕 - "B站搜一下 AI 教程" →
bili search "AI 教程" - "全网搜 LLM 框架对比" → Exa 语义搜索
- "订阅这个 RSS" →
feedparser
Agent 读取 SKILL.md 后自己知道该调什么。
安装的系统依赖
install --system 会安装以下外部工具(--safe/默认模式只检查、不安装):
| 工具 | 用途 |
|---|---|
| gh CLI | GitHub 读取/搜索 |
| Node.js | yt-dlp JS runtime + mcporter |
| mcporter + Exa | 全网语义搜索 |
| bili-cli | B站搜索/详情(可选) |
安全性
| 措施 | 说明 |
|---|---|
| 🔒 无凭据 | 本版本不需要 Cookie/Token,配置仅存一个可选的网络代理 |
| 🛡️ 默认安全 | install 默认只检查不改系统;显式 --system 才安装工具 |
| 👀 完全开源 | 代码透明,随时可审查 |
| 🔍 Dry Run | install --dry-run 预览所有操作,不做任何改动 |
| 🧩 可插拔 | 不信任某个组件?换掉对应 channel 文件即可 |
命令参考
agent-reach install --env=auto [--system | --safe] [--dry-run] [--proxy=...]
agent-reach doctor [--json]
agent-reach version
License
致谢
yt-dlp · Jina Reader · feedparser · gh CLI · bili-cli · Exa · mcporter
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.