Two tools with minimal documentation. interactive_feedback has a basic description (Chinese/English) but lacks parameter descriptions, the schema shows 'feedback', 'command_logs', and 'images' fields with only type and description keys, but no guidance on format, constraints, or when to use each. get_system_info has no parameters but its description is vague ('Get system environment information'). Neither tool documents output schemas, error handling, or recovery paths. No tool annotations (readOnlyHint, destructiveHint, idempotentHint). Naming is acceptable (verb_noun pattern), but descriptions fall short of the 50 - 200 char LLM-optimized baseline and lack actionable context.

• Add a detailed description to get_system_info (50 - 200 chars): 'Returns the current system environment details (OS, Python version, installed packages, memory, CPU). Call this first to verify the runtime environment before executing system-dependent tasks.'
• Document all parameters of interactive_feedback with type, format, and constraints: feedback (string, max 5000 chars, required), command_logs (string, optional, format: newline-separated log lines), images (array of base64-encoded image strings, optional, max 5 images).
• Add output schema documentation for both tools. For get_system_info, specify returned fields (os_name, os_version, python_version, installed_packages, memory_mb, cpu_count). For interactive_feedback, specify success response (status, feedback_id, timestamp).
• Implement error handling with recovery guidance: 'If images fail to upload, try reducing image size or count. If command_logs are too large, truncate to last 1000 lines.' Return structured error responses with actionable next steps.
• Add tool annotations: mark interactive_feedback with destructiveHint=true (writes feedback to system), mark get_system_info with readOnlyHint=true (no side effects).

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

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