Files
temporal-ai-agent/tools/search_flights.py
Steve Androulakis 5d55a9fe80 Model Context Protocol (MCP) support with new use case (#42)
* initial mcp

* food ordering with mcp

* prompt eng

* splitting out goals and updating docs

* a diff so I can get tests from codex

* a diff so I can get tests from codex

* oops, missing files

* tests, file formatting

* readme and setup updates

* setup.md link fixes

* readme change

* readme change

* readme change

* stripe food setup script

* single agent mode default

* prompt engineering for better multi agent performance

* performance should be greatly improved

* Update goals/finance.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* Update activities/tool_activities.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* co-pilot PR suggested this change, and now fixed it

* stronger wording around json format response

* formatting

* moved docs to dir

* moved image assets under docs

* cleanup env example, stripe guidance

* cleanup

---------

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
2025-06-09 16:39:57 -07:00

341 lines
11 KiB
Python

import http.client
import json
import os
import random
import urllib.parse
from dotenv import load_dotenv
def search_airport(query: str) -> list:
"""
Returns a list of matching airports/cities from sky-scrapper's searchAirport endpoint.
"""
load_dotenv(override=True)
api_key = os.getenv("RAPIDAPI_KEY", "YOUR_DEFAULT_KEY")
api_host = os.getenv("RAPIDAPI_HOST_FLIGHTS", "sky-scrapper.p.rapidapi.com")
conn = http.client.HTTPSConnection(api_host)
headers = {
"x-rapidapi-key": api_key,
"x-rapidapi-host": api_host,
}
# Sanitize the query to ensure it is URL-safe
print(f"Searching for: {query}")
encoded_query = urllib.parse.quote(query)
path = f"/api/v1/flights/searchAirport?query={encoded_query}&locale=en-US"
conn.request("GET", path, headers=headers)
res = conn.getresponse()
if res.status != 200:
print(f"Error: API responded with status code {res.status}")
print(f"Response: {res.read().decode('utf-8')}")
return []
data = res.read()
conn.close()
try:
return json.loads(data).get("data", [])
except json.JSONDecodeError:
print("Error: Failed to decode JSON response")
print(f"Response: {data.decode('utf-8')}")
return []
def search_flights_real_api(
args: dict,
) -> dict: # rename to search_flights to use the real API
"""
1) Looks up airport/city codes via search_airport.
2) Finds the first matching skyId/entityId for both origin & destination.
3) Calls the flight search endpoint with those codes.
"""
date_depart = args.get("dateDepart")
date_return = args.get("dateReturn")
origin_query = args.get("origin")
dest_query = args.get("destination")
# Step 1: Resolve skyIds
origin_candidates = search_airport(origin_query)
destination_candidates = search_airport(dest_query)
if not origin_candidates or not destination_candidates:
return {"error": "No matches found for origin/destination"}
origin_params = origin_candidates[0]["navigation"]["relevantFlightParams"]
dest_params = destination_candidates[0]["navigation"]["relevantFlightParams"]
origin_sky_id = origin_params["skyId"] # e.g. "LOND"
origin_entity_id = origin_params["entityId"] # e.g. "27544008"
dest_sky_id = dest_params["skyId"] # e.g. "NYCA"
dest_entity_id = dest_params["entityId"] # e.g. "27537542"
# Step 2: Call flight search with resolved codes
load_dotenv(override=True)
api_key = os.getenv("RAPIDAPI_KEY", "YOUR_DEFAULT_KEY")
api_host = os.getenv("RAPIDAPI_HOST_FLIGHTS", "sky-scrapper.p.rapidapi.com")
conn = http.client.HTTPSConnection(api_host)
headers = {
"x-rapidapi-key": api_key,
"x-rapidapi-host": api_host,
}
path = (
"/api/v2/flights/searchFlights?"
f"originSkyId={origin_sky_id}"
f"&destinationSkyId={dest_sky_id}"
f"&originEntityId={origin_entity_id}"
f"&destinationEntityId={dest_entity_id}"
f"&date={date_depart}"
f"&returnDate={date_return}"
f"&cabinClass=economy&adults=1&sortBy=best&currency=USD"
f"&market=en-US&countryCode=US"
)
conn.request("GET", path, headers=headers)
res = conn.getresponse()
data = res.read()
conn.close()
try:
json_data = json.loads(data)
except json.JSONDecodeError:
return {"error": "Invalid JSON response"}
itineraries = json_data.get("data", {}).get("itineraries", [])
if not itineraries:
return json_data # Return raw response for debugging if itineraries are empty
formatted_results = []
seen_carriers = set()
for itinerary in itineraries:
legs = itinerary.get("legs", [])
if len(legs) >= 2:
# Extract outbound and return flight details
outbound_leg = legs[0]
return_leg = legs[1]
# Get the first segment for flight details
outbound_flight = outbound_leg.get("segments", [{}])[0]
return_flight = return_leg.get("segments", [{}])[0]
# Extract flight details
outbound_carrier = outbound_flight.get("operatingCarrier", {}).get(
"name", "N/A"
)
outbound_carrier_code = outbound_flight.get("operatingCarrier", {}).get(
"alternateId", ""
)
outbound_flight_number = outbound_flight.get("flightNumber", "N/A")
outbound_flight_code = (
f"{outbound_carrier_code}{outbound_flight_number}"
if outbound_carrier_code
else outbound_flight_number
)
return_carrier = return_flight.get("operatingCarrier", {}).get(
"name", "N/A"
)
return_carrier_code = return_flight.get("operatingCarrier", {}).get(
"alternateId", ""
)
return_flight_number = return_flight.get("flightNumber", "N/A")
return_flight_code = (
f"{return_carrier_code}{return_flight_number}"
if return_carrier_code
else return_flight_number
)
# Check if carrier is unique
if outbound_carrier not in seen_carriers:
seen_carriers.add(outbound_carrier) # Add to seen carriers
formatted_results.append(
{
"outbound_flight_code": outbound_flight_code,
"operating_carrier": outbound_carrier,
"return_flight_code": return_flight_code,
"return_operating_carrier": return_carrier,
"price": itinerary.get("price", {}).get("raw", 0.0),
}
)
# Stop after finding 3 unique carriers
if len(formatted_results) >= 3:
break
return {
"origin": origin_query,
"destination": dest_query,
"currency": "USD",
"results": formatted_results,
}
def generate_smart_flights(origin: str, destination: str) -> list:
"""
Generate realistic flight options with smart pricing based on origin and destination.
"""
# Common airlines for different regions
airlines_by_region = {
"domestic_us": [
{"name": "American Airlines", "code": "AA"},
{"name": "United Airlines", "code": "UA"},
{"name": "Delta Airlines", "code": "DL"},
{"name": "Southwest Airlines", "code": "WN"},
],
"us_international": [
{"name": "American Airlines", "code": "AA"},
{"name": "United Airlines", "code": "UA"},
{"name": "Delta Airlines", "code": "DL"},
{"name": "Virgin Atlantic", "code": "VS"},
],
"australia_nz": [
{"name": "Qantas", "code": "QF"},
{"name": "Jetstar", "code": "JQ"},
{"name": "Virgin Australia", "code": "VA"},
{"name": "Air New Zealand", "code": "NZ"},
],
"international": [
{"name": "American Airlines", "code": "AA"},
{"name": "United Airlines", "code": "UA"},
{"name": "Delta Airlines", "code": "DL"},
{"name": "Air New Zealand", "code": "NZ"},
{"name": "Qantas", "code": "QF"},
{"name": "Singapore Airlines", "code": "SQ"},
],
}
# Determine route type and base pricing
origin_lower = origin.lower()
dest_lower = destination.lower()
# Australia/NZ cities
anz_cities = [
"sydney",
"melbourne",
"syd",
"mel",
"auckland",
"akl",
"wellington",
"wlg",
"brisbane",
"bne",
"perth",
"per",
]
# US cities
us_cities = [
"los angeles",
"lax",
"san francisco",
"sfo",
"new york",
"nyc",
"jfk",
"chicago",
"ord",
"miami",
"mia",
]
is_origin_anz = any(city in origin_lower for city in anz_cities)
is_dest_anz = any(city in dest_lower for city in anz_cities)
is_origin_us = any(city in origin_lower for city in us_cities)
is_dest_us = any(city in dest_lower for city in us_cities)
# Determine airline pool and base price
if (is_origin_us and is_dest_anz) or (is_origin_anz and is_dest_us):
# Trans-Pacific routes
airline_pool = airlines_by_region["international"]
base_price = random.randint(1200, 1800)
elif is_origin_anz and is_dest_anz:
# Australia/NZ domestic
airline_pool = airlines_by_region["australia_nz"]
base_price = random.randint(300, 600)
elif is_origin_us and is_dest_us:
# US domestic
airline_pool = airlines_by_region["domestic_us"]
base_price = random.randint(200, 800)
else:
# General international
airline_pool = airlines_by_region["international"]
base_price = random.randint(800, 1500)
# Generate 3-4 flight options
num_flights = random.randint(3, 4)
results = []
used_airlines = set()
for i in range(num_flights):
# Pick unique airline
available_airlines = [a for a in airline_pool if a["name"] not in used_airlines]
if not available_airlines:
available_airlines = airline_pool # Reset if we run out
airline = random.choice(available_airlines)
used_airlines.add(airline["name"])
# Generate flight numbers
outbound_num = random.randint(100, 999)
return_num = random.randint(100, 999)
# Price variation (cheaper airlines get lower prices)
price_multiplier = 1.0
if "Southwest" in airline["name"] or "Jetstar" in airline["name"]:
price_multiplier = 0.7
elif "Virgin" in airline["name"]:
price_multiplier = 0.85
elif "Singapore" in airline["name"]:
price_multiplier = 1.2
# Add some random variation
price_variation = random.uniform(0.9, 1.1)
final_price = round(base_price * price_multiplier * price_variation, 2)
results.append(
{
"operating_carrier": airline["name"],
"outbound_flight_code": f"{airline['code']}{outbound_num}",
"price": final_price,
"return_flight_code": f"{airline['code']}{return_num}",
"return_operating_carrier": airline["name"],
}
)
# Sort by price
results.sort(key=lambda x: x["price"])
return results
def search_flights(args: dict) -> dict:
"""
Search for flights. Uses real API if RAPIDAPI_KEY is available, otherwise generates smart mock data.
"""
load_dotenv(override=True)
api_key = os.getenv("RAPIDAPI_KEY")
origin = args.get("origin")
destination = args.get("destination")
if not origin or not destination:
return {"error": "Both origin and destination are required"}
# If API key is available, use the real API
if api_key and api_key != "YOUR_DEFAULT_KEY":
return search_flights_real_api(args)
# Otherwise, generate smart mock data
results = generate_smart_flights(origin, destination)
return {
"currency": "USD",
"destination": destination,
"origin": origin,
"results": results,
}