Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. search_knowledge_base has a reasonable description (89 chars) and two typed parameters with descriptions. get_document has a minimal description (54 chars) and one parameter. Neither tool documents output schema structure, which is critical for LLM planning. No error handling guidance, no parameter constraints (e.g., top_k bounds), and no indication of what fields are returned. Naming is clear (verb_noun pattern), but descriptions lack context on when to use each tool vs. the other, and do not explain the citation/relevance score structure returned.

• Document output schema for both tools. For search_knowledge_base, specify: [{doc_id: string, title: string, source: string, content: string, relevance_score: float}]. For get_document, specify: {doc_id: string, title: string, source: string, content: string, metadata: object}.
• Add parameter constraints: top_k should have min=1, max=50 (or appropriate limit). Update description: 'Number of top results to return (1-50, default 3).'
• Enhance search_knowledge_base description to explain when to use it: 'Search the knowledge base for relevant passages. Use this to find documents matching a query. Returns ranked results with relevance scores. If you have a specific doc_id, use get_document instead.'
• Enhance get_document description: 'Retrieve the full text of a document by its ID. Use this after search_knowledge_base returns a doc_id you want to read in full. Returns complete document with metadata.'
• Add error handling guidance: 'If doc_id is not found, the search_knowledge_base tool can help locate the correct ID. If the knowledge base is empty, no results will be returned.'

Inferred effective spec: <=2025-11-25.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry