# skill-seekers/Skill_Seekers [Health: Active]

**Category:** 🧠 Knowledge & Memory  
**Repository:** https://github.com/yusufkaraaslan/Skill_Seekers  
**GitHub Stars:** 14953  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/skill-seekers-skill-seekers

## Description
Transform 17 source types (docs, GitHub repos, PDFs, videos, Jupyter, Confluence, Notion, Slack/Discord) into AI-ready skills and RAG knowledge. 35 MCP tools for scraping, packaging, enhancing, and exporting to vector databases (Weaviate, Chroma, FAISS, Qdrant). Supports 16+ target platforms.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `uvx` (confidence: high):

```json
"mcpServers": {
  "skill-seekers": {
    "command": "uvx",
    "args": ["skill-seekers"]
  }
}
```

## Documentation

## What skill-seekers/Skill_Seekers MCP server does

skill-seekers/Skill_Seekers MCP server prepares source material for AI systems. It can collect information from documentation sites, GitHub repositories, local codebases, PDFs, Word documents, EPUB books, Jupyter notebooks, OpenAPI files, PowerPoint presentations, AsciiDoc, HTML, RSS or Atom feeds, and man pages. Optional support covers videos, Confluence, Notion, and Slack or Discord chat exports.

The resulting assets can be packaged as AI skills containing a `SKILL.md` and reference files, converted into chunked documents with metadata for retrieval workflows, or formatted for vector databases such as Pinecone, Chroma, Weaviate, FAISS, and Qdrant. Export targets include AI platforms and coding assistants such as Claude, Gemini, OpenAI, Cursor, Windsurf, and Cline.

## How it works

The primary workflow is source creation followed by optional enhancement and packaging. A source can be supplied as a URL, repository identifier, local path, or source-specific command options. For example, the CLI can create a skill from a documentation URL, a GitHub repository, a local project, or a PDF. The output can then be packaged for a selected target.

For project analysis, the `scan` command examines manifests, README files, Docker or CI configuration, and sampled source imports. It produces configurations for detected frameworks and a codebase configuration that can be used in a later creation step. AI enhancement can use the default Claude agent, another named agent such as Kimi, or a custom agent command.

skill-seekers/Skill_Seekers MCP server adds MCP access to these scraping, scanning, packaging, enhancement, and export workflows. The README identifies the MCP installation extra for clients including Claude Code, Cursor, and other MCP-compatible environments.

## Setup and configuration

Python 3.10 or newer and Git are listed as prerequisites. Install the base package for core scraping, repository handling, PDF processing, and packaging:

```bash
pip install skill-seekers
```

Add the MCP extra to install MCP support:

```bash
pip install 'skill-seekers[mcp]'
```

Optional extras cover all LLM integrations, video processing, notebooks, PowerPoint, Confluence, Notion, RSS, chat exports, AsciiDoc, or the complete feature set. Video transcript and metadata extraction uses the `video` extra; visual extraction and Whisper support use `video-full`, followed by the project’s setup command for GPU-aware dependencies.

## Tools and capabilities

- Create structured skills from web pages, repositories, local projects, and many file formats.
- Process Confluence, Notion, and Slack or Discord exports when the corresponding extras are installed.
- Scan codebases and generate framework-specific configuration files.
- Enhance collected content with a selected or custom AI agent.
- Package outputs for AI platforms and coding assistants.
- Prepare chunked RAG content and vector-database-ready data.
- Export material for LangChain, LlamaIndex, Haystack, Pinecone, Chroma, Weaviate, FAISS, and Qdrant workflows.

## Limitations and notes

Feature availability depends on the relevant optional extra. Video workflows, for example, require separate video dependencies, while source systems such as Confluence, Notion, and chat exports have their own extras. The README describes both 18 source types and an MCP badge showing 40 tools, while the supplied listing describes 17 source types and 35 MCP tools; the exact totals should therefore be checked against the installed release.

skill-seekers/Skill_Seekers MCP server is primarily a preparation and export layer. The provided material does not indicate that it hosts a vector database, supplies an embedding service, or provides a managed remote MCP endpoint. External services and AI agents may require their own credentials and configuration.

_Full upstream README: https://allmcps.com/mcp/skill-seekers-skill-seekers/readme_