I built a self-hosted AI knowledge base on top of offline Wikipedia (and it speaks MCP)

Description and use

Zim Pro replaces kiwix‑serve, enabling full‑text search, AI querying, automatic updates, and offline LLM integration.

Problem

The existing kiwix-serve tool only allows reading offline ZIM files; the author needed to query the content, keep it updated automatically, and use it as a source for local LLMs.

Outcome

Zim Pro enables offline querying of large ZIM libraries with AI-generated answers, automatic updates, and integration with local LLMs, all without internet access.

Workflow

• 1. Install Zim Pro via Docker or pip.
• 2. Mount ZIM files into the /data volume or add them via the built-in catalog.
• 3. Start the server and access http://localhost:8083 to browse and search.
• 4. Use the MCP command claude mcp add --transport http zim-pro http://localhost:8083/mcp to expose tools to Claude.
• 5. Query the library with AI tools; Zim Pro performs plan, deep search, light search, and rank steps to retrieve relevant passages.
• 6. The AI answers are repaired for Markdown correctness before display.

Evidence

On my home server I have a folder containing ZIM files: these are offline versions of Wikipedia, Stack Exchange, DevDocs and iFixit, all drawn from the Kiwix catalog.

Source

The problem was that kiwix-serve only allows you to read it. What I wanted to do was: ask it questions and get answers I could verify, keep the books updated without checking for new releases by hand, have my other AI tools (Claude, the local LLMs) use it as a source.

Source

Zim Pro makes use of the full-text index that is already contained in each ZIM file and places the intelligence around that index: Plan, Deep search, Light search, Rank.

Source

The available tools are list_books, search, suggest, get_article, get_article_outline and list_updates.

Source