- Home
- Skills
- Research & Analysis
- Offline Wikipedia MCP Setup (Kiwix ZIM)
More screenshots
Works with the AI Agents you already use
Offline Wikipedia MCP Setup (Kiwix ZIM)
Give your local agent good research habits for offline Wikipedia.
$1
Offline Wikipedia MCP Setup (Kiwix ZIM)
Example session with this skill installed
Set up the offline Wikipedia MCP server for me in LM Studio. I've downloaded the Wikipedia Mini nopic .zim file.
- Read your context and instructions
- Compiled the offline wikipedia mcp
Offline Wikipedia MCP: setup complete ✅
I installed the server, configured LM Studio, and verified it works.
What I did
| Step | Result |
|---|---|
| Checked Python | 3.12 ✅ (3.10+ required) |
| Created venv + installed deps | libzim and mcp installed |
| Found your archives | 1 file in ~/kiwix-zim/ |
| Wrote LM Studio config | Developer → mcp.json updated |
| Test call | search_wikipedia("Alan Turing") returned results ✅ |
Config I added
{
"mcpServers": {
"offline-wikipedia": {
"command": "/Users/you/mcp-offline-wikipedia/venv/bin/python",
"args": ["/Users/you/mcp-offline-wikipedia/mcp_wiki_server.py"],
"env": {
"WIKI_ZIM_DIR": "/Users/you/kiwix-zim/",
"WIKI_MAX_CHARS": "15000"
}
}
}
}
Things I avoided for you
- Your .zim files are in
~/kiwix-zim/, not~/Downloads, so macOS won't block access. - I used the venv Python, not system
python3.
Next steps
- Restart LM Studio and enable the
offline-wikipediaserver. - Ask your model: "Use Wikipedia to explain the Turing test."
- Want more coverage? Drop the Maxi archive into the same folder and it loads automatically.
Connects securely to your tools. The creator never sees your data.
About this skill
Give your local agent good research habits for offline Wikipedia.
The Offline Wikipedia MCP server lets agents read local Kiwix .zim archives, but agents often use it badly: they ask full questions instead of keywords, pass the [archive label] into get_article, or flood the context window. This skill fixes that.
What it teaches your agent
- Two-step retrieval:
search_wikipediafirst to find the exact title, thenget_article - Query formulation: turn "who was the first US president?" into
George Washington - Clean title contract: strip the number prefix and
[archive]label before fetching - Context safety: read 1–2 articles per turn and handle
WIKI_MAX_CHARStruncation - Zero-result recovery: broaden, use synonyms, try alternate spellings
- Bundled extras: a
verify_setup.pypre-flight check, plus MCP config and Kiwix archive guides
Requirements
You need the free, open-source MCP server already installed: github.com/Rome-NotBeepBoop/mcp-offline-wikipedia. This skill does not install it; it makes your agent better at using it.
Requires local .zim files and libzim. Knowledge is limited to the date of the archive snapshot. Cannot search local source code.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 14 days ago
Requires
Permissions
File Scopes
Network is used only for git clone and pip install during setup. A timestamped backup is created before any client config is modified. The server itself runs fully offline.