8.3
/ 10
1 evaluations
1.7k Downloads
Overview
Provide flight search and destination discovery via the Jinko MCP server, enabling the assistant to find routes, prices, and travel ideas based on user-specified origins, budgets, dates, and preferences.
Key Advantages
1.Two focused tools (find_destination and find_flight) that cover both open-ended trip ideas and concrete route searches.
2.Clear, detailed parameter schema (origins, destination lists, date ranges, stay lengths, budget, cabin, direct-only, currency, locale, sorting).
3.Strong guidance on airport and city code handling, including multi-airport cities (NYC, LON, PAR, TYO, etc.) and thematic destination sets (beach, ski, warm in winter, etc.).
4.Designed to be proactive: encourages filling as many parameters as possible and re-calling the tools when user criteria change.
5.Works well as a general trigger for most flight-related intents: searching flights, cheap deals, trip planning, and destination discovery from a given origin.
Use Cases
- User wants ideas on where to go from a given city within a budget and time window (e.g., “Somewhere warm from Chicago in December for a week”).
- User wants to compare flights for a specific route with flexible dates or stay lengths (e.g., “NYC to London, 5–10 days in business class”).
- User is looking for cheap flights from an origin to a region or themed destination set (e.g., “SF to Europe under $800”, “beach destinations from BOS”).
- User wants to explore the cheapest dates to a single destination (e.g., “Cheapest dates from LAX to Tokyo in November”).
- User is generally planning a trip and wants deal discovery from their home airport (“Find good flight deals from Miami over the next few months”).
Evaluation Scores
8.3
/ 10
Reliability
7.0
Functionality
8.5
Usability
9.0
Safety
9.0
Performance
7.5
Compatibility
8.0
Based on 1 evaluation · Latest: 3/19/2026
Download Trend
Loading...
Evaluation History (1)
8.3/103/19/2026▼
OS: linux-x64LLM: anthropic/claude-sonnet-4.6
**Judgment:** A well-designed, high-utility flight search and destination-discovery skill that is especially strong for travel planning, deal hunting, and “where should I go?” scenarios. Documentation is detailed and prescriptive, which makes correct tool usage straightforward and reduces integration errors.
**Strengths**
- Covers both open-ended destination discovery (`find_destination`) and specific route search (`find_flight`).
- Rich parameter set (dates, date ranges, stay lengths, budget, cabin class, direct-only, currency, locale, sorting) enables tailored searches.
- Clear rules for interpreting user intent (when to use which tool, default to roundtrip, only use oneway when explicit, all dates must be future).
- Strong guidance on mapping cities to multiple airports and generating destination IATA codes from thematic criteria (beach, ski, warm in winter, regional groupings).
**Key Risks / Limitations**
- Hard dependency on the external Jinko MCP server (`https://mcp.gojinko.com`); outages or network issues will break functionality.
- Flight data (availability, prices) may be time-sensitive and can diverge from what users see on booking sites; should not be treated as a guaranteed booking quote.
- Requires careful date handling: the assistant must ensure all dates are in the future and properly parse “next month”, “in December”, “weekend”, etc.
- City/airport code expansion and region-to-IATA mapping rely on the assistant following the documented examples; incorrect mappings can degrade results.
**Recommended Scenarios**
- Use as the primary tool whenever users ask about flights, cheap deals, or destination ideas from a specific origin.
- Use `find_destination` when the user is exploring and has not locked in a destination (or only has vague criteria like climate, activities, or regions).
- Use `find_flight` when both origin and destination are known, including when the user wants the cheapest dates for that route.
- Avoid positioning it as a booking engine; treat it as a search and discovery assistant whose results should be verified on actual booking platforms.
Comments (0)
No comments yet. Be the first!