Three tools with complete input schemas and descriptions. Naming follows verb_noun pattern (search_, get_, list_). Descriptions are adequate (100-200 chars) but lack actionable context for LLM selection. Parameters are typed and mostly constrained via enums. Critical gap: no output schemas documented, LLMs cannot plan downstream calls or extract structured data. Error handling and recovery guidance absent. Market auto-detection is clever but undocumented in tool descriptions.

• Document output schemas for all three tools. Example for search_property_comps: {type: 'object', properties: {sales: [{address, price, date, property_type, area}], statistics: {median, average, min, max}}}. This enables LLMs to extract and chain results.
• Expand tool descriptions to include WHEN/WHY. E.g., 'Use search_property_comps to find recent comparable sales in a specific area. Use get_area_stats for aggregate market statistics without individual listings.'
• Add explicit error handling: 'If location is not recognized, try list_markets() to see supported areas, or provide a postcode/ZIP code instead of a named area.'
• Constrain property_type as an enum per market, or accept a free-form string and validate server-side with a clear error message: 'Invalid property_type for market {market}. Valid types: {list}.'
• Document the market auto-detection heuristics in the 'market' parameter description so LLMs understand when to omit it vs. when to specify explicitly.