Two tools with minimal definitions. list_skills has a basic description (54 chars) but no input schema. read_skill has a description (56 chars) and one parameter with type and description, but lacks output schema documentation. Neither tool has error handling guidance, recovery paths, or pagination support. No tool annotations (readOnlyHint, etc.). Naming is acceptable (verb_noun pattern), but descriptions are below the 194-char baseline for production tools. No evidence of idempotency guarantees, rate limiting, or security considerations.

• Add explicit input schema to list_skills with optional parameters: limit (1-100, default 20), offset (default 0), and filter (optional string). Document pagination in description.
• Document output schema for both tools. list_skills should return {skills: [{name: string, description: string, category?: string}], total: number, hasMore: boolean}. read_skill should return {name: string, description: string, content: string, metadata?: object}.
• Expand descriptions to 150-250 chars. Example for list_skills: 'Retrieve all available skills from the canonical store. Returns paginated results with skill names and summaries. Use this to discover available skills before calling read_skill to fetch full details.'
• Add error handling guidance. E.g., 'If a skill is not found, the tool returns a 404 with available alternatives. If the store is unavailable, retry after 30 seconds.'
• Add tool annotations via MCP: mark both tools with readOnlyHint=true (they do not modify state). This helps agents reason about safety and retry behavior.