Server exposes 4 read-only tools with basic descriptions and partial schemas. Tool names follow verb_noun convention (get_*, list_, search_) which is good. However, parameter descriptions are minimal or absent, output schemas are not documented, and error handling lacks recovery guidance. The server implements caching and MediaWiki integration well, but the MCP tool definitions themselves are underdeveloped for production agent use.

• Document the output schema for each tool. For get_page_content, specify: {url: string, title: string, description: string, markdown: string, lastModified?: string, metadata: object}. For list_content and get_sitemap, specify the array structure and field types.
• Add min/max bounds to search_site's 'limit' parameter. E.g., 'Maximum number of results to return (1 - 100, default 20)'. This prevents LLMs from requesting unreasonable result sets.
• Expand tool descriptions to 100 - 200 characters. For get_sitemap vs list_content, clarify the distinction: 'get_sitemap returns a hierarchical XML-like structure; list_content returns a flat array of all pages with metadata.' This helps LLMs choose the right tool.
• Add error recovery guidance to get_page_content description: 'If the page is not found, try search_site() with keywords from the page title.' This guides agents on next steps.
• Document pagination for list_content and get_sitemap if they support it. If results are capped at MW_MAX_PAGES (100), state this in the description and explain how to iterate through results.