This MCP server has critical gaps in definition quality. Both tools have minimal descriptions (under 50 chars), lack parameter descriptions entirely, and have no visible output schema documentation. The input schemas are present but sparse. Naming is acceptable but descriptions fail to meet baseline standards for LLM-optimized tool selection. Per the rubric baseline of 194 chars average for tool descriptions and 100% of A+ tools having param descriptions, this server falls significantly short.

• Expand tool descriptions to 150-250 chars. For search: 'Search the web for information related to a query using Tavily. Returns a list of relevant results with titles, URLs, and snippets. Use when you need current information not in your training data.'
• Expand web_scrapping description: 'Scrape and extract content from a URL using Firecrawl. Returns markdown and HTML versions of the page content. Use when you need to read or extract specific information from a webpage.'
• Add parameter descriptions: query → 'The search query as a natural language string (e.g., "latest AI news"). Aim for 1-10 words for best results.'; url → 'A valid HTTP or HTTPS URL (e.g., https://example.com/article). Must be publicly accessible.'
• Document output schemas explicitly: search → 'Returns dict with keys: results (list of {title, url, snippet, score}), answer (str), raw_answer (str)'; web_scrapping → 'Returns ScrapeResponse with markdown (str), html (str), metadata (dict)'
• Add error handling in docstrings: 'If the Tavily API fails (quota exceeded, invalid key), returns an error explaining the failure. Suggest checking TAVILY_API_KEY or retrying after a delay.'
• Rename web_scrapping to scrape_url for clarity (verb_noun pattern). Rename search to search_web to disambiguate from other possible search tools.
• Add constraints: query parameter should have max length (e.g., 500 chars); url should be validated to start with http:// or https://
• Include when-to-use guidance in descriptions: search should say 'when you need current information or facts not in your training data'; web_scrapping should say 'when you need to extract or read specific text from a webpage'.