KFabric's 7 tools have minimal descriptions and incomplete parameter documentation. Tool names follow verb_noun convention (discover_, collect_, analyze_, consolidate_, generate_, build_, prepare_) but descriptions are absent or trivial in the source. Input schemas are visible with basic type info (string, array) but lack enums, ranges, or validation constraints. No output schemas documented. Error handling is not evident. The server exposes domain-specific operations but lacks the LLM-optimized descriptions and parameter guidance required for reliable agent selection and invocation.

• Add 50 - 200 character descriptions to all 7 tools explaining WHAT they do, WHEN to call them, and WHAT they return. E.g., 'discover_documents: Find candidate documents matching a query. Call this first to identify relevant sources before analyzing. Returns list of document IDs and titles.'
• Document every parameter with type, format, and constraints. E.g., 'query_id (string, 1 - 64 chars): Unique identifier for the search query. Obtain from create_query or list_queries.'
• Define output schemas for each tool. Specify returned fields, types, and structure. E.g., discover_documents returns {documents: [{id: string, title: string, score: number}], total: number}.
• Add error handling guidance for WRITE tools. Document what can fail, how to recover, and whether operations are idempotent. E.g., 'collect_candidate: If document already collected, returns success (idempotent). If query_id invalid, returns 404 with available query IDs.'
• Add parameter enums or patterns where applicable. E.g., if fragment_ids must be UUIDs, state 'fragment_ids: array of UUID strings (36 chars, format: xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx)'.