This server has critical definition quality gaps. Tool descriptions are present but minimal (10-50 chars, well below the 50-200 char baseline for LLM-optimized descriptions). Input schemas are visible for most tools, but many lack depth, parameters often have minimal descriptions. The 18 tools span backup/restore, credential configuration, service management, and security scanning, but descriptions do not explain WHEN to use each tool, WHAT it returns, or dependencies between them. No output schemas are documented. Error handling is absent from descriptions. The server exposes secrets as parameters (tokens, API keys) in configure_* tools, violating the secret-injection pattern. Tool naming is reasonable (verb_noun style) but descriptions fail to guide LLM selection.

• Move all credential parameters (token, api_key, app_id, app_secret, corp_id, agent_id, secret, webhook_url) to server-side environment variables or a secure vault. Remove them from tool parameters entirely. Update configure_* tools to accept only a 'confirm' flag and read credentials from stdin or a secure prompt.
• Expand all tool descriptions to 80-150 characters. Include: (1) what the tool does, (2) when to use it vs similar tools, (3) what it returns. Example: 'backup' → 'Create a compressed, optionally encrypted backup of All in One Claw config, credentials, and workspace. Returns backup file path. Use before major updates or when migrating systems.'
• Document output schemas for all tools. For 'backup', specify: {backup_file: string, size: string, encrypted: boolean}. For 'list', specify: {backups: [{filename: string, size: string, created_at: string}], total: number}. For 'verify_gateway', specify: {running: boolean, port: number, status: string}.
• Add error handling guidance to descriptions. E.g., restore: 'If backup file is corrupted, restore will fail. Ensure file is valid and not truncated. A safety backup is created before restore, check /tmp/ if restore fails.' For fix_* tools: 'If fix fails, check service logs with systemctl status openclaw-gateway.'
• Add parameter descriptions for enums. E.g., scan_skill's 'format' param: 'Output format. Use "text" for human-readable report, "json" for programmatic parsing by downstream tools.'