borski/travel-hacking-toolkit

What it solves

This toolkit provides AI agents with the ability to perform complex "travel hacking"—optimizing travel bookings using points, miles, and award flights. It automates the search for the best value across dozens of mileage programs, compares cash prices against point redemptions, and manages loyalty balances to find the most cost-effective way to fly or stay in hotels.

How it works

The project implements a suite of Model Context Protocol (MCP) servers and specialized "skills" that can be integrated into AI tools like Claude Code, Codex, and OpenCode. It leverages a mix of official APIs (such as Seats.aero and Duffel) and browser-automated tools (via Patchright and Docker containers) to scrape data from sites without public APIs (e.g., Southwest, American Airlines, and credit card portals).

Who it’s for

Frequent travelers and "travel hackers" who use AI agents to automate the tedious process of searching for award availability, calculating transfer ratios, and planning multi-city trips.

Highlights

• Award Flight Search: Accesses award availability across 27 different mileage programs.
• Unified Comparison: Parallel search across multiple sources for flights and hotels to find the best price or point value.
• Unified Loyalty Management: Integrates with AwardWallet and direct account checks to track points and elite status.
• Comprehensive Reference Knowledge: Includes built-in institutional knowledge on airline alliances, stopover rules, and point valuations.
• Detailed Transit Integration: Includes specialized tools for European rail and ferry routes (e.g., Deutsche Bahn, Scandinavia Transit).
• Browser Automation: Uses Docker-based skills to access gated portals like Amex Travel and Chase Travel.