This server has 8 tools with reasonable naming (all start with action verbs) and descriptions present for most tools. However, there are significant gaps in parameter descriptions, schema completeness, and error handling guidance. The tool definitions are explicitly visible in src/qq_agent_mcp/tools.py and src/qq_agent_mcp/onebot.py, so no inference penalty applies. Descriptions average ~100-150 characters, which is acceptable but some lack context on WHEN to use the tool or what prerequisites exist. Parameter descriptions are present but inconsistent in detail, some parameters have good descriptions (e.g., target_type with enum and clear text), while others are sparse. No output schemas are documented, and error handling provides no recovery guidance to the LLM. The tool set addresses a specific domain (QQ messaging via NapCat) with reasonable separation of concerns, but falls short of production-grade quality due to missing output documentation and incomplete parameter guidance.

• Document the output schema for each tool. Use JSON Schema or a clear textual description of fields, types, and structure. For example, get_recent_messages should document that it returns an array of message objects with fields: message_id (string), sender_id (string), sender_name (string), content (string), timestamp (ISO 8601 string), etc.
• Add pagination to list_groups and list_friends: include limit (default 20, max 100) and offset/cursor parameters. Return a response with total_count, items array, and next_cursor (if more results exist). This prevents context window exhaustion on large contact lists.
• Expand tool descriptions to 80-150 characters, explaining WHEN and WHY to call the tool. Example for list_groups: 'Retrieve all groups the bot has joined. Call this first to discover available group targets before sending a message to a group.'
• Add error handling guidance to descriptions of tools that can fail (send_message, get_recent_messages, get_group_info, screenshot_chat). Example: 'If target_id is invalid, the tool returns an error. Try list_groups() or list_friends() to find valid targets.'
• Document interdependencies between optional parameters in send_message. Specify that num_chunks overrides split_content if both are provided, and that both are mutually exclusive with the </分段> tag splitting.
• For screenshot_chat, add a note about prerequisites: 'Requires the Playwright browser automation engine to be initialized. If screenshot fails, check that the QQ web interface is accessible and Playwright is properly configured.'

Inferred effective spec: <=2025-11-25.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry