Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and output structure. query_document has a reasonable description (89 chars) and schema with two parameters, but lacks output schema documentation. get_collection_info has minimal description (47 chars) and no visible parameters or output schema. Neither tool includes error handling guidance, recovery hints, or structured output documentation. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) present. Parameter descriptions are present but generic. No pagination support visible despite semantic search returning multiple results. Baseline: 549 production tools average 194 chars for descriptions and 100% have documented return types, this server falls short on both counts.

• Document output schema for query_document: return array of {score: float, page: int, content: string} with total_count and next_cursor for pagination support.
• Expand get_collection_info description to 100+ chars explaining what metadata is returned and when to call it (e.g., 'Call before search to verify collection is loaded and see document count').
• Add min/max constraints to num_results: 'Number of results to return (1 - 50, default: 5). Higher values increase latency and token usage.'
• Add error handling guidance to both tools: 'If ChromaDB connection fails, verify OPENAI_API_KEY is set and ./chroma_db path exists. Call get_collection_info to diagnose.'
• Add readOnlyHint=true annotation to both tools in Tool registration to signal safety and enable agent optimization.

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

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