README.md
Secure MCP server for MediaWiki wikis.
mcp-mediawiki-crunchtools
Secure MCP server for MediaWiki wikis. Search, read, create, edit, and manage wiki pages, categories, files, and more. Works with any MediaWiki instance — public or private.
Installation
uvx (recommended)
uvx mcp-mediawiki-crunchtools
pip
pip install mcp-mediawiki-crunchtools
mcp-mediawiki-crunchtools
Container
podman run -e MEDIAWIKI_URL=https://en.wikipedia.org/w/ \
quay.io/crunchtools/mcp-mediawiki
Usage with Claude Code
Read-only (public wiki, no auth needed)
claude mcp add mcp-mediawiki-crunchtools \
--env MEDIAWIKI_URL=https://en.wikipedia.org/w/ \
-- uvx mcp-mediawiki-crunchtools
With authentication (for write operations)
claude mcp add mcp-mediawiki-crunchtools \
--env MEDIAWIKI_URL=https://your-wiki.com/w/ \
--env MEDIAWIKI_USERNAME=BotUser \
--env MEDIAWIKI_PASSWORD=BotPassword \
-- uvx mcp-mediawiki-crunchtools
HTTP transport (systemd / container)
podman run -d --name mcp-mediawiki \
-p 127.0.0.1:8016:8016 \
-e MEDIAWIKI_URL=https://your-wiki.com/w/ \
quay.io/crunchtools/mcp-mediawiki \
--transport streamable-http --host 0.0.0.0
Claude Code config:
{"type": "http", "url": "http://127.0.0.1:8016/mcp"}
Environment Variables
| Variable | Required | Description |
|---|---|---|
MEDIAWIKI_URL |
Yes | Wiki base URL (e.g., https://en.wikipedia.org/w/) |
MEDIAWIKI_USERNAME |
No | Bot/user account for write operations |
MEDIAWIKI_PASSWORD |
No | Bot/user password |
MEDIAWIKI_HTTP_USER |
No | HTTP Basic Auth username (.htaccess) |
MEDIAWIKI_HTTP_PASS |
No | HTTP Basic Auth password |
Tools (19)
| Category | Tool | Description |
|---|---|---|
| Pages | search |
Full-text search across wiki |
| Pages | get_page |
Get page wikitext content |
| Pages | get_page_html |
Parse page to HTML |
| Pages | list_pages |
List pages with prefix filter |
| Pages | create_page |
Create a new page |
| Pages | edit_page |
Edit an existing page |
| Pages | delete_page |
Delete a page |
| Pages | move_page |
Move/rename a page |
| Categories | list_categories |
List all categories |
| Categories | get_category_members |
Get pages in a category |
| Categories | get_page_categories |
Get categories for a page |
| Recent Changes | list_recent_changes |
List recent edits |
| Parsing | parse_wikitext |
Parse raw wikitext to HTML |
| Site Info | get_site_info |
Get wiki config and version |
| Site Info | list_namespaces |
List wiki namespaces |
| Users | get_user_info |
Get user details |
| Users | list_user_contributions |
List user edits |
| Files | get_file_info |
Get file/image metadata |
| Files | list_files |
List files on the wiki |
Development
# Install dev dependencies
uv sync --all-extras
# Quality gates
uv run ruff check src tests
uv run mypy src
uv run pytest -v
gourmand --full .
podman build -f Containerfile .
License
AGPL-3.0-or-later
Tools (19)
searchFull-text search across wikiget_pageGet page wikitext contentget_page_htmlParse page to HTMLlist_pagesList pages with prefix filtercreate_pageCreate a new pageedit_pageEdit an existing pagedelete_pageDelete a pagemove_pageMove/rename a pagelist_categoriesList all categoriesget_category_membersGet pages in a categoryget_page_categoriesGet categories for a pagelist_recent_changesList recent editsparse_wikitextParse raw wikitext to HTMLget_site_infoGet wiki config and versionlist_namespacesList wiki namespacesget_user_infoGet user detailslist_user_contributionsList user editsget_file_infoGet file/image metadatalist_filesList files on the wikiEnvironment Variables
MEDIAWIKI_URLrequiredWiki base URLMEDIAWIKI_USERNAMEBot/user account for write operationsMEDIAWIKI_PASSWORDBot/user passwordMEDIAWIKI_HTTP_USERHTTP Basic Auth usernameMEDIAWIKI_HTTP_PASSHTTP Basic Auth passwordConfiguration
claude_desktop_config.json
{"type": "http", "url": "http://127.0.0.1:8016/mcp"}Try it
→Search the wiki for pages related to 'Project Documentation' and summarize the findings.
→Create a new wiki page titled 'Meeting Notes' with the content provided in my current context.
→List all pages currently in the 'Technical' category.
→Get the recent changes from the wiki and identify any edits made by the user 'BotUser'.
→Parse the wikitext of the 'Home' page and convert it to HTML for me.