Single tool 'trigger_doc_update' has critical definition gaps. Description is in Chinese (non-English), lacks clarity on prerequisites, return value, and when to use it. Input schema is empty object {} with no parameters documented. No output schema defined. Tool performs destructive write operation (WRITE risk) but lacks confirmation pattern or dry-run capability. No error handling guidance. Naming is acceptable (verb_noun form) but description fails LLM optimization baseline (194 chars avg; this is ~120 chars in Chinese, unclear when translated).

• Translate description to English and expand to 100 - 200 chars. Example: 'Triggers a background task to fetch and update documentation from the official website. Overwrites existing Markdown files in data/ directory. Returns job ID, file count, and completion timestamp. Requires write permissions. May take 30 - 120 seconds.'
• Define input schema with optional parameters: e.g., 'force_refresh' (boolean, default false), 'timeout_seconds' (integer, 30 - 600, default 120). Document each parameter's purpose and constraints.
• Add output schema documenting: job_id (string), status (enum: queued|running|completed|failed), files_updated (integer), timestamp (ISO 8601), error_message (string, nullable). Example: {"job_id": "job_abc123", "status": "completed", "files_updated": 42, "timestamp": "2025-01-15T10:30:00Z", "error_message": null}
• Implement confirmation pattern: add a 'confirm' boolean parameter (default false). When false, return a dry-run summary (files to be overwritten, estimated size). When true, execute the update. Prevents accidental data loss.
• Add error handling guidance in description: 'If network fails, retry after 30 seconds. If disk is full, free space and retry. If permission denied, check file ownership. Contact admin if errors persist.'

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

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