Critical gaps across all dimensions. Tool definitions lack formal schemas, parameter descriptions are minimal or absent, and descriptions are generic. Only 2 of 5 tools have documented input parameters; 3 tools have no visible schema definitions. Descriptions are under 100 chars for most tools, providing minimal LLM guidance. No error handling patterns, no output schema documentation, and no security annotations. This is a bash-based skill collection, not a production-grade MCP server.

• Add formal JSON Schema input definitions for all tools. At minimum: type, description, and enum constraints for each parameter. Example for bytestash-api command: {"type": "string", "enum": ["list", "search", "get", "create", "push", "update", "delete", "share", "shares", "unshare", "view-share"], "description": "API command to execute"}.
• Expand tool descriptions to 150-250 chars. Include: (1) what it does, (2) when to use it vs similar tools, (3) what it returns. Example: 'Fetch all PR review comments and threads from GitHub using GraphQL. Call this first to see what feedback exists before addressing comments. Returns array of comment objects with thread IDs, author, and resolution status.'
• Document output schemas for each tool. Specify field names, types, and which fields are required for chaining. Example for fetch_comments: returns {comments: [{id, author, body, thread_id, resolved}], total_count}.
• Add error handling guidance in descriptions. Example: 'If authentication fails, verify QBIT_URL and QBIT_KEY in ~/.claude-homelab/.env. If rate-limited, retry after 60 seconds.'
• Implement tool annotations (readOnlyHint, destructiveHint, idempotentHint) for protocol compliance. Mark qbit-api and bytestash-api write operations as destructiveHint=true; mark fetch_comments and verify_resolution as readOnlyHint=true.