Server provides 11 tools with basic descriptions and input schemas. All tools follow verb_noun naming (search_, get_, list_, recommend_, validate_, suggest_, compare_). Descriptions are present but generic (avg ~80 chars, below 194-char baseline). Input schemas visible in code but lack detail: most parameters are bare strings with minimal constraint info. No output schemas documented. No error handling guidance. All tools are READ_ONLY, reducing risk, but composition and chaining support is weak, tools do not clearly return IDs needed for downstream calls. Parameter descriptions are minimal (e.g., 'Search query to find matching workflow specs' lacks format/length constraints). No pagination support despite list tools. Missing tool annotations (readOnlyHint present implicitly via risk=READ_ONLY, but not in schema).

• Add output schema documentation to all tools. Example: 'Returns: {spec_name: string, category: string, description: string, steps: array[string], triggers: array[string]}'
• Enhance parameter descriptions with constraints. Change 'Search query to find matching workflow specs' to 'Search query (1-100 chars, alphanumeric + spaces). Searches name, description, and tags. Returns up to 20 results.'
• Add pagination to list_specs_by_category and list_all_specs: limit (1-100, default 20), offset (default 0), and return total_count.
• Document error cases and recovery paths. E.g., 'If spec_name not found, call list_all_specs to see available specs, then retry with correct name.'
• Add tool annotations to schema: mark all tools with readOnlyHint=true since all are READ_ONLY. This enables client-side filtering and safety checks.