OmniSeek

io.github.Battam1111Published by GitHub account Battam1111v0.2.1Updated Oct 24, 2026

Your agent seeks what search can't find. A self-hosted perception MCP server.

VerifiedSTDIODesktop onlyWeb Search & ScrapingKnowledge & Memory

Overview

AI-generated overview

OmniSeek is a self-hosted MCP server that lets an AI assistant search sources web search misses, such as Bilibili, Zhihu, V2EX and research papers.

What you can do
Search Chinese community sources such as Bilibili videos and comments, Zhihu answers and recent V2EX threads, with links to the original posts, and cross languages by asking in English and searching in Chinese or the reverse. Find the papers a research field is built on and recent papers citing them. Save a search once and rerun it to see only what is new since last time. Optionally transcribe podcasts and videos to text locally and quote them with timestamps, and read PDFs. Keep findings in a local index so later searches start from what it already knows, and let the agent keep notes through its ruling and statement tools.
When to use it
Useful when an assistant needs material that ordinary web search does not surface, especially Chinese-language communities, exact quotes from videos or podcasts, or citation trails in research papers. Also fits recurring topic monitoring where only new results matter.
Requirements
Runs locally as a stdio process started by the AI app; install with uv tool install and Python 3.11 or newer (uv can install it). No account, API key or open port is required. Optional add-ons cover podcast transcription, PDF reading and logged-in sites. Network access to the enabled sources is needed.
Before you install
Sites that require a login are off by default; when switched on, the agent reads them through the login already in your browser, so consider which accounts you expose. Searches go directly to the sites you enable, and results are kept in local files. Optional transcription and PDF add-ons must be installed separately.

Installation

In SourceWeft

  1. Open OmniSeek in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

[OmniSeek]

Give your AI agent the parts of the internet web search doesn't reach.

OmniSeek is an add-on for AI assistants and agents like Claude Code, Cursor, and Codex that lets them search where web search stops.

  • Ask in English, get what people actually said on Bilibili, and Zhihu when you ask for it, with links to the original posts.
  • Get the papers a research field is built on, and the recent ones building on them, in one answer.
  • Have your agent watch a topic and tell you only what is new since last time.
bash
uv tool install omniseekclaude mcp add omniseek -- omniseek

Runs on your own computer. It plugs in through MCP, the standard way AI apps like Claude Code and Cursor add tools. Cursor and Codex: one line each, see Install.

[CI] [PyPI] [License] [Python] [Built for MCP] Languages: English · 中文 · 日本語

See the difference

We asked two agents for Mu Li's exact words on the Transformer paper's writing, and the minute, in his 87-minute Bilibili walkthrough: web search returned a paraphrase and "around minutes 1 to 8", while the agent with OmniSeek transcribed the video and found the line at 1:21:51.

[A real run, side by side. Left, an agent with web search only answers with a paraphrase and says around minutes 1 to 8. Right, an agent with OmniSeek transcribes the 87-minute video and quotes Mu Li at 1:21:51, with a link that opens there.] [A real run, side by side. Left, an agent with web search only answers with a paraphrase and says around minutes 1 to 8. Right, an agent with OmniSeek transcribes the 87-minute video and quotes Mu Li at 1:21:51, with a link that opens there.]

The full comparison on three questions, with sources, dates, and a short video.

What your agent can reach

  • Chinese communities: Bilibili videos and comments, Zhihu answers when you ask for them, recent V2EX threads, all with links to the posts. More: Bilibili, Zhihu
  • Across languages: ask in English, search in Chinese, answer in English. And the reverse. More
  • Podcasts and videos: turns speech into text on your own computer and quotes it with the minute (optional add-on). More
  • Research papers: which papers a field is built on and which recent ones cite them. More
  • New since last time: save a search once; each rerun shows only what is new. More
  • Sites you log in to: if you switch it on, your agent reads them through your own browser login. Off by default. More

It also keeps what it finds in a local index on your computer, so later searches start from what it already knows.

Install

The two commands above are all you need, with Python 3.11 or newer (uv installs it if you don't have it). Your AI app starts OmniSeek when it needs it. Nothing to keep running, no port to open.

Cursor, Codex, pipx, Docker, and the optional add-ons (podcast transcription, PDF reading, sites you log in to) are in Install.

Your data, your logins

OmniSeek runs on your computer. There is no OmniSeek cloud, no account, no telemetry: your agent's searches go straight to the sites you turned on. Public sites are on by default. Sites that need a login are off until you switch them on, and then they use the login already in your own browser; OmniSeek never sees your password. What it builds up as you use it stays in local files you own.

OmniSeek has no AI model of its own and sends nothing to a server of ours. Your agent does the thinking; OmniSeek goes and gets things. Settings: Configuration.

Tools

Your agent starts with omniseek_search. The rest, from paper maps to saved searches and the notes your agent keeps (omniseek_ruling, omniseek_statement), are in tools.md.

More

Worked examples, real outputs · A full case study · FAQ, including how it differs from Exa and Tavily · All sources · Source health, updated weekly

How we test these claims: a test only counts if we first checked that plain web search could not find its answer on the first page (latest results). Using Claude Code? skills/omniseek-investigate is a ready-made research method: search wide, narrow in, add structure.

Contributing

See CONTRIBUTING.md: a new source has to find something plain web search doesn't. Fixing a source that stopped working is always welcome. Run python tests/smoke.py before you push.

By participating you agree to the Code of Conduct.

Give your AI agent the parts of the internet web search doesn't reach.

Apache-2.0 · NOTICE · Security · Cite

Source: README.md at commit b3b1293

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v0.2.1LatestSep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts