Two tools with basic verb-noun naming (search_*, get_*) but significant quality gaps. search_fuel_status has a minimal 1-field schema with no parameter description. get_fuel_summary has an empty schema. Both descriptions are present but generic (under 100 chars). No output schemas documented. No error handling guidance. No parameter validation hints. Tool logic is simple read-only operations, but definitions lack the rigor expected for production use. Schemas are visible in mcp_server.py but lack depth; descriptions do not explain WHEN to use each tool or what structure is returned.

• Add detailed parameter descriptions: 'query (string): Search term matching station name, province, or district. E.g., "Bangkok", "PTT", "Chatuchak". Case-insensitive partial match.'
• Document output schema as structured JSON: search_fuel_status should return {"count": int, "results": [{"stationName": string, "province": string, "diesel": string, "benzine95": string, "confidence": float}]}
• Add error recovery hints: 'If no results found, try a broader search term or check data availability via get_fuel_summary.'
• Expand tool descriptions to 100-200 chars explaining WHEN to use each: 'search_fuel_status: Find fuel availability at specific stations or regions. Use this to check if a particular area has diesel or benzine shortages.'
• Implement pagination for search_fuel_status: add optional 'limit' (1-100, default 20) and 'offset' parameters; return total count so LLM knows if more results exist.