A multi-agent travel planning system that integrates aviation, weather, and hotel search via MCP servers to generate comprehensive travel itineraries
Server has 5 tools with minimal quality. Tool names follow verb_noun pattern (list_*, get_*), which is good. However, descriptions are generic and lack actionable context. Input schemas are present but sparse, most tools have only basic type info without constraints, ranges, or validation guidance. No output schemas documented. Error handling is absent; tools will fail silently or return raw API errors. Parameters lack descriptions in most cases. No pagination, no result limits documented. Security: API keys are hardcoded in environment variables (acceptable) but no secret injection pattern documented. Overall: definitions are skeletal, suitable only for trivial use cases.
Fetches current weather information for a specified city including temperature, humidity, wind speed, and weather condition
Fetches a 5-entry weather forecast for a specified city with datetime, temperature, and weather conditions
Returns a list of major international airlines with their codes, names, and countries
Returns a list of major international airports with their codes, names, cities, and countries
Searches the web for information using Tavily's search API, returns results with titles, URLs, and content snippets
No output schemas documented. LLMs cannot plan downstream tool calls or extract structured data. Tools return raw API responses without field documentation.
Input parameter 'city' in get_current_weather and get_forecast lacks description. LLMs cannot infer whether to pass city name, code, or coordinates.
No error handling or recovery guidance. Tools will fail on invalid city names, API errors, or network timeouts with no actionable message for the LLM.
No pagination or result limits documented. list_airports and list_airlines return fixed datasets, but no guidance on handling large result sets or pagination.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 47 | 2026-07-28+ | v2 |
Tool descriptions are generic and lack context for LLM selection. E.g., 'Returns a list of major international airports' does not explain WHEN to call it or what the LLM should do with the result.