Three tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. All tools have names starting with action verbs (get_, google_) and descriptions present, but descriptions are minimal (34-56 chars, below the 50-200 char baseline for LLM optimization). Parameter descriptions exist but lack detail on constraints, formats, and valid ranges. Output schemas are not documented, LLMs cannot predict response structure. Error handling returns text blobs rather than structured guidance. No tool annotations (readOnlyHint, etc.). Tools are read-only and low-risk, but definition quality is below production baseline.

• Expand tool descriptions to 100-150 chars. Example: 'Retrieve content of a webpage in markdown format. Use this to extract full page text, articles, or documentation. Requires a valid URL and optional wait time for dynamic content loading.'
• Document output schemas explicitly. For google_search, specify: 'Returns a JSON object with fields: results (array of {title, url, snippet}), total_results (number), search_time_ms (number).'
• Add constraint details to parameter descriptions. For 'country', write: 'ISO 3166-1 alpha-2 country code (e.g., US, GB, CA). Defaults to US if omitted.'
• Implement structured error responses with recovery guidance. Instead of 'Experience API Error: 401', return: {isError: true, errorCode: 'AUTH_FAILED', message: 'API key is invalid or expired.', recovery: 'Check EXPERIENCE_API_KEY environment variable in claude_desktop_config.json.'}
• Add tool annotations to schema. For all three tools, add readOnlyHint: true to signal they are safe to call repeatedly without side effects.