Server has 10 tools with complete input schemas and basic descriptions. All tools follow verb_noun naming (search_*, get_*). However, descriptions are minimal (avg ~60 chars, well below the 194-char baseline for A+ tools) and lack context on WHEN to use each tool or dependencies between them. No output schemas documented. Parameters lack format constraints, ranges, or validation guidance. Error handling is absent, no recovery guidance or categorization. Tools are read-only but lack explicit idempotent hints. Dependency injection pattern (WeiboCrawler) is clean but not visible in tool definitions themselves.

• Expand tool descriptions to 100-200 chars. Include WHEN to use (e.g., 'Call this after search_users() to fetch detailed profile info'), what it returns, and any prerequisites. Example: 'get_profile: Fetch detailed user profile including follower count, verification status, and bio. Requires a valid uid from search_users() or get_followers().'
• Document output schemas explicitly. Update tool return types from generic list[dict] to reference the Pydantic models in schemas.py, or inline the field descriptions in docstrings. E.g., 'Returns: list[FeedItem] with fields: id, text, created_at, comments_count, attitudes_count, reposts_count.'
• Add parameter constraints. Use JSON Schema minValue/maxValue for limit (1-100) and page (1+). Add descriptions like 'limit: Number of results to return (1-100, default 15)' and 'page: Page number starting from 1.'
• Implement error handling with recovery guidance. Catch common failures (invalid uid, no results, rate limit) and return structured errors like '{"error": "User not found", "suggestion": "Try search_users() with a partial name first."}'
• Add tool annotations via FastMCP. Mark all tools with readOnlyHint=true and idempotentHint=true so LLMs know these are safe to retry and do not modify state.