Flight MCPblueprint

AI flight booking, solved

MCPrates Flight MCP gives your AI agent real-time AI flight search, live AI flight prices, and fare-rule intelligence — so it can quote, compare and rebook flights the way a senior travel agent would, in seconds.

AI flight bookingAI flight searchAI flight pricesFare rules parsingPrice-drop alertsNDC + GDS

The agent's reality

Why booking flights hurts

Fare rules are unreadable

Fare basis, penalties, advance-purchase, min/max stay — buried in cryptic GDS text. Agents read them by hand, and get them wrong.

Prices move every minute

You quote a fare, the client hesitates, and by the time they say yes it has jumped $120. Re-shopping manually is a chore.

Every GDS speaks differently

Amadeus, Sabre, Travelport and NDC each have their own shape. Normalizing them eats your day.

Complex itineraries take forever

Multi-city, open-jaw, mixed-cabin trips mean dozens of manual searches and a spreadsheet to compare.

Schedule changes blindside you

Airlines move flights; you find out from an angry client. Catching it early is manual monitoring.

Baggage & seats are guesswork

What's included? What's the change fee? It's scattered across fare rules and ancillary tables.

Show, don't tell

What your AI does with flights now

Parse Fare Rules

tool_call

"Is this $482 Emirates fare refundable, and what's the change fee?"

parse_fare_rules(EK029)

{ "fareBasis": "QLOWUAE",
  "refundable": false,
  "changeable": true,
  "changeFee": "GBP 150",
  "baggage": "30kg",
  "minStay": "SUN"
}

Read in 200ms · no human decoding

Recheck Price Drops

tool_call

"Watch this quoted fare and tell me if it drops below $450."

watch_fare(threshold=450)

09:00  $482  holding
09:30  $471
10:05  $448  ✓ ALERT
→ "rebook": true, save $34

Rebooked before the client asked

Live Fare Search

tool_call

"Cheapest direct DXB→LHR next Tuesday under $500."

search_flights()

[ "Emirates EK029" $482 0 stops
  "BA106"        $496 0 stops
  "Virgin VS401" $511 ✗ over
]  3 providers · live

Normalized across GDS + NDC

Complex Itineraries

tool_call

"Build a DXB→ROM→BCN→DXB multi-city in business under $3k."

search_flights(multiCity)

leg1 DXB→FCO EK097 $1,140
leg2 FCO→BCN VY6011 $610
leg3 BCN→DXB EK186 $1,090
total $2,840 ✓ under budget

Stitched automatically

Change Eligibility

tool_call

"Client wants to move to Thursday — is it allowed and how much?"

parse_fare_rules + reprice()

{ "changeAllowed": true,
  "penalty": "GBP 150",
  "fareDiff": "GBP 64",
  "total": "GBP 214"
}

Answer in one turn

Schedule-Change Watch

tool_call

"Alert me if any of my ticketed PNRs get a schedule change."

watch_schedule(PNR*)

⚠ PNR X9F2KQ EK097
  was 02:15 → now 04:40
  "connectionRisk": true,
  "action": "reaccommodate"

Caught before the client noticed

Flight AI questions, answered

Can AI book flights?

Yes. Through the MCPrates Flight MCP server, an AI agent (Claude, ChatGPT, Cursor and others) can search live flights, read fare rules, price an itinerary and hand off a ready-to-ticket booking — all via standard MCP tool calls.

How does AI find the cheapest flights?

The agent calls search_flights with your origin, destination and dates; MCPrates queries multiple GDS and NDC providers in real time and returns normalized fares so the AI can compare price, stops and cabin instantly.

Can AI read and explain fare rules?

Yes. The parse_fare_rules tool turns raw fare-basis text into structured fields — refundable, changeable, penalties, baggage and stay requirements — so your agent answers "is this refundable?" without a human decoding it.

Can AI track flight price drops?

Yes. watch_fare re-polls a quoted fare on an interval and alerts the agent the moment the price drops below a threshold, so you can rebook the same itinerary cheaper before the client even asks.

Give your agent flight superpowers

Live fares, fare rules and price-drop alerts — one MCP endpoint.

Request Access