8 tools with basic schemas and descriptions, but significant gaps. All tools are READ_ONLY and follow verb_noun naming (get_*, search_*). Descriptions are present but minimal (10-60 chars), below the 194-char baseline. No output schemas documented. No error handling guidance. Parameters like 'size' and 'contact_id' have descriptions, but many lack context on valid ranges, formats, or dependencies. Tool composition is reasonable (search + get pattern), but descriptions don't explain WHEN to use each tool or what to expect in responses.

• Expand tool descriptions to 100-200 chars. Include WHAT the tool does, WHEN to use it (vs similar tools), and what data it returns. Example: 'Search for contacts by name or email. Returns up to [size] results with contact_id, name, email. Use this to find a contact before calling get_contact for full details.'
• Document output schemas for all tools. Specify which fields are returned (contact_id, name, email, etc.), their types, and which IDs chain to downstream tools. This enables LLM planning.
• Add min/max constraints to 'size' parameter. E.g., 'Number of results per page (1-100, default 20)'. Document pagination behavior: does it return total_count? next_cursor? Are results sorted?
• Add error recovery hints to get_* tools. E.g., 'Get a contact by ID (UUID). Returns contact details or error if not found, try search_contacts() first if you only have a name.'
• Clarify tool composition in descriptions. E.g., search_contacts description should say 'Use this to discover contacts; then call get_contact(contact_id) for full details.' This guides LLM planning.

Inferred effective spec: 2025-06-18+.

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