Autosearch
installs/mo · pypistarsforks
How to install Autosearch?
0xmariowu/autosearch
$git clone https://github.com/0xmariowu/AutosearchOpen-source deep research for AI agents: 40 channels, 10+ Chinese sources.
README.md
| 1 | <div align="center"> |
| 2 | |
| 3 | # AutoSearch |
| 4 | |
| 5 | **Open-source Deep Research for AI Agents** |
| 6 | |
| 7 | *40 channels, including 10+ Chinese sources.*<br> |
| 8 | *MCP-native. LLM-decoupled. Plug into the agent host you already use.* |
| 9 | |
| 10 | [](https://github.com/0xmariowu/Autosearch/actions/workflows/ci.yml) |
| 11 | [](https://github.com/0xmariowu/Autosearch/releases) |
| 12 | [](https://www.npmjs.com/package/autosearch-ai) |
| 13 | [](LICENSE) |
| 14 | [](https://github.com/0xmariowu/Autosearch) |
| 15 | [](https://modelcontextprotocol.io) |
| 16 | |
| 17 | [Install](#install) · [Channels](docs/channels.mdx) · [MCP Setup](docs/mcp-clients.md) · [Examples](docs/) · [Docs](https://docs.autosearch.dev) · [中文](README.zh.md) |
| 18 | |
| 19 | </div> |
| 20 | |
| 21 | AutoSearch is open-source deep research infrastructure built for AI agents. Plug Claude Code, Cursor, Cline, GPT-Researcher, LangChain, LlamaIndex, AutoGen, and other hosts into MCP-native access across 40 channels, including 10+ Chinese sources. |
| 22 | |
| 23 | The engine returns indexed multi-source results and stays uncoupled from LLM calls, so your agent keeps its own model, prompts, and workflow. |
| 24 | |
| 25 | --- |
| 26 | |
| 27 | You ask your AI to research something. It answers from training data cutoff — |
| 28 | |
| 29 | - "Show me this week's LLM papers on arxiv" → **can't**, no academic database access |
| 30 | - "What are people saying about this product on Reddit" → **shallow**, only surface-level web results |
| 31 | - "Find similar open-source projects on GitHub" → **weak**, general search misses most repos |
| 32 | - "Summarize the Twitter discussion on this topic" → **blocked**, no public API |
| 33 | - "Compare opinions on Hacker News vs Chinese tech forums" → two platforms, manual aggregation |
| 34 | |
| 35 | **AutoSearch fixes this in one line.** Pick the path that matches you: |
| 36 | |
| 37 | ## Install |
| 38 | |
| 39 | **Have Node?** (most common — works on macOS, Linux, Windows) |
| 40 | |
| 41 | ```bash |
| 42 | npx autosearch-ai |
| 43 | ``` |
| 44 | |
| 45 | **Using Claude Code / Cursor / Zed?** Paste this into your AI agent: |
| 46 | |
| 47 | ``` |
| 48 | Help me install AutoSearch: https://raw.githubusercontent.com/0xmariowu/Autosearch/main/docs/install.md |
| 49 | ``` |
| 50 | |
| 51 | **Shell user on macOS / Linux?** |
| 52 | |
| 53 | ```bash |
| 54 | curl -fsSL https://raw.githubusercontent.com/0xmariowu/Autosearch/main/scripts/install.sh | bash |
| 55 | ``` |
| 56 | |
| 57 | The npm wrapper runs install/init only when you invoke it explicitly. Plain |
| 58 | `npm install -g autosearch-ai` does not auto-run `init`; avoiding npm |
| 59 | install-time scripts is intentional supply-chain hardening. |
| 60 | |
| 61 | After install, your Agent searches 40 channels simultaneously — academic papers, developer communities, Chinese social media, and Linux DO forum threads — results deduplicated and ranked, every result includes a source URL. |
| 62 | |
| 63 | --- |
| 64 | |
| 65 | ## Channels |
| 66 | |
| 67 | | Channel | Ready out of box | Unlocked after config | How to configure | |
| 68 | |---------|-----------------|----------------------|-----------------| |
| 69 | | 📄 **arxiv** | Academic preprints CS/ML/Physics | — | No config needed | |
| 70 | | 🔬 **PubMed / OpenAlex / DBLP** | Biomedical + cross-discipline papers | — | No config needed | |
| 71 | | 🐙 **GitHub** | Code, Issues, repository search | — | No config needed | |
| 72 | | 🔍 **DuckDuckGo** | General web search | — | No config needed | |
| 73 | | 🟠 **Hacker News** | Developer discussions + early product signals | — | No config needed | |
| 74 | | 📖 **Reddit** | Community discussion + real user experiences | — | No config needed | |
| 75 | | 🤗 **Hugging Face** | Open-source ML model search | — | No config needed | |
| 76 | | 📦 **Stack Overflow** | Programming Q&A | — | No config needed | |
| 77 | | 📰 **Google News** | Real-time news aggregation | — | No config needed | |
| 78 | | 💻 **dev.to** | Developer blogs | — | No config needed | |
| 79 | | 📊 **SEC EDGAR** | US company filings | — | No config needed | |
| 80 | | 🌐 **Wikipedia / Wikidata** | Encyclopedia + structured knowledge | — | No config needed | |
| 81 | | 💬 **WeChat Official Accounts** | Full-text article search | — | No config needed | |
| 82 | | 📱 **36kr / InfoQ** | Chinese tech & business news | — | No config needed | |
| 83 | | 🧵 **Linux DO / Discourse Forums** | Linux DO topics and Discourse-based public forum discussions | — | No config needed | |
| 84 | | 🎬 **YouTube** | Video search + transcripts | — | Set `YOUTUBE_API_KEY` | |
| 85 | | 📺 **Bilibili** | Chinese tech videos | — | Set `TIKHUB_API_KEY` | |
| 86 | | 📹 **WeChat Channels** | Chinese short videos (视频号) | — | `TIKHUB_API_KEY` | |
| 87 | | 🌸 **Xiaohongshu** | Lifestyle + product reviews | More stable | `autosearch login xhs` or `TIKHUB_API_KEY` | |
| 88 | | 🐦 **Twitter / X** | Real-time discussion + tech news | — | `TIKHUB_API_KEY` | |
| 89 | | 📣 **Weibo / Douyin / Zhihu** | Chinese social + in-depth discussion | — | `TIKHUB_API_KEY` | |
| 90 | | 📈 **Xueqiu** | A-share / HK stock discussion | — | `autosearch login xueqiu` | |
| 91 | | 💼 **LinkedIn** | Public pages via Jina Reader | — | No config needed | |
People who liked this also viewed
6 picksType
- ddgsA metasearch library that aggregates results from diverse web search servicesMCP ServersMay 2026
- firecrawl-mcp-server🔥 Official Firecrawl MCP Server - Adds powerful web scraping and search to Cursor, Claude and any other LLM clients.MCP ServersJul 2026
- fliGoogle Flights MCP, CLI and Python LibraryMCP ServersJul 2026
- sembleFast and Accurate Code Search for Agents. Uses ~98% fewer tokens than grep+readMCP ServersJul 2026
- duckduckgo-mcp-serverA Model Context Protocol (MCP) server that provides web search capabilities through DuckDuckGo, with additional features for content fetching and parsing.MCP ServersJul 2026
- wikipedia-mcpA Model Context Protocol (MCP) server that retrieves information from Wikipedia to provide context to LLMs.MCP ServersMay 2026
