Flight Delay Prediction

API version: 1.0.6

Auth

No authentication required.

Base URL

https://test.api.amadeus.com/v1

Setup

1. No auth setup needed
2. GET /travel/predictions/flight-delay -- verify access

Endpoints

1 endpoints across 1 groups. See references/api-spec.lap for full details.

travel

Method	Path	Description
GET	/travel/predictions/flight-delay	Return the delay segment where the flight is likely to lay.

Enhanced Skill Content

Question Mapping

• "Will my flight be delayed?" -> GET /travel/predictions/flight-delay
• "What is the delay probability for flight AA123?" -> GET /travel/predictions/flight-delay
• "How likely is a delay flying from JFK to LAX tomorrow?" -> GET /travel/predictions/flight-delay
• "Predict delay for a 3-hour flight on Boeing 737" -> GET /travel/predictions/flight-delay
• "What are the chances my evening departure gets delayed?" -> GET /travel/predictions/flight-delay
• "Is carrier DL likely to be on time for flight 456?" -> GET /travel/predictions/flight-delay
• "Compare delay risk for morning vs afternoon departures" -> GET /travel/predictions/flight-delay (call twice, vary departureTime)
• "Which arrival time has the lowest delay probability?" -> GET /travel/predictions/flight-delay (call multiple times, compare results)
• "How does aircraft type affect delay prediction for this route?" -> GET /travel/predictions/flight-delay (call with different aircraftCode values)
• "Get delay forecast for a short-haul domestic flight" -> GET /travel/predictions/flight-delay
• "What delay category is most likely for my redeye flight?" -> GET /travel/predictions/flight-delay
• "Should I expect my connecting flight to be delayed?" -> GET /travel/predictions/flight-delay (call per leg)

Response Tips

• Predictions: Response contains probability buckets by delay category (e.g., on-time, 15-30 min, 30-60 min, 60+ min) -- surface the highest-probability category first and present all tiers so the user can assess risk.
• Errors (400): Bad request typically means a missing or malformed required parameter -- all 10 parameters are mandatory, so check each before retrying.

Anomaly Flags

• All parameters required: If any of the 10 required fields is missing, the call will fail with 400. Pre-validate before sending.
• Date/time mismatches: Flag if arrivalDate is before departureDate, or arrivalTime implies a negative duration relative to departureTime (accounting for timezone differences).
• Duration inconsistency: Surface a warning if the supplied duration does not roughly match the time delta between departure and arrival.
• Unusual aircraft codes: Flag if aircraftCode is not a recognized IATA aircraft type (e.g., typos like "7X7" instead of "737").
• Test vs production base URL: The spec uses test.api.amadeus.com -- remind the user this is the sandbox environment, not live data.
• Auth token expiry: Amadeus APIs use OAuth2 tokens that expire quickly (typically 1799s) -- surface a reminder to refresh if calls start returning 401.

Playbook

1. Predict Delay for a Known Flight

1. Gather flight details: carrier code, flight number, departure/arrival airports, dates, times, aircraft type, and duration.
2. Format dates as YYYY-MM-DD and times as HH:MM:SS.
3. Call GET /travel/predictions/flight-delay with all 10 required parameters.
4. Read the prediction result and identify the highest-probability delay bucket.
5. Present the full probability distribution to the user with the most likely outcome highlighted.

2. Compare Delay Risk Across Multiple Flights

1. Collect details for each candidate flight (e.g., different carriers or times on the same route).
2. Call GET /travel/predictions/flight-delay once per flight option.
3. Extract the on-time probability from each response.
4. Rank flights from lowest to highest delay risk.
5. Present a comparison table showing each flight's delay probability breakdown.

3. Assess Delay Risk for a Multi-Leg Itinerary

1. Break the itinerary into individual flight legs.
2. Call GET /travel/predictions/flight-delay for each leg with its specific parameters.
3. Identify any legs with high delay probability (e.g., >40% chance of 30+ min delay).
4. Flag legs where a delay could cause a missed connection based on layover time.
5. Summarize overall itinerary risk and recommend buffer time for tight connections.

4. Evaluate Time-of-Day Impact on Delays

1. Pick a fixed route, carrier, flight number, and aircraft type.
2. Call GET /travel/predictions/flight-delay varying departureTime and arrivalTime across morning, afternoon, and evening slots (adjust duration if needed).
3. Collect the on-time probability for each time slot.
4. Present results showing which departure window has the best on-time performance.
5. Recommend the optimal departure window to the user.

Response Tips

• Check response schemas in references/api-spec.lap for field details

References

• Full spec: See references/api-spec.lap for complete endpoint details, parameter tables, and response schemas

Generated from the official API spec by LAP