TourMind Travel Booking Skills
Search, compare, and book hotels and flights worldwide through a conversational agent.
- Source repo
- tourmind-com/Tourmind-Booking-Skills
- Stars
- ★ 2.1k
- Last updated
- 9d ago
- License
- MIT
- Primary language
- Python
- FA score
- 56/100 · Major gaps
At a glance
- How it runs
- Works with
- Universal · cross-platformChatGPT · Codex · Claude Code
- Setup effort
- Low · running in minutes
- You'll need
- Typical use
- A leisure traveler planning a Tokyo stay wants to compare live room rates, room types, breakfast, and cancellation terms before choosing.
- Not a fit if
- Users who need to search or book entirely offline
- Travelers who need flight cancellation or change actions
- Source review
- 56/100 · Major gaps
What does this agent do, and when should you use it?
This repository provides hotel and flight skills for personal and business users, used through natural language in supported Skills-enabled agent clients. It includes tourmind-booking for hotels and flight-booking-ai for flights, using the TourMind API to retrieve live inventory, room rates, and flight offers. Users can compare options, recheck final prices and availability, then follow the conversation workflow to place bookings, look up orders, and handle supported payments or hotel cancellations. Searching does not create an order; bookings, payments, and eligible hotel cancellations require explicit user confirmation. Network access is required; hotel and flight search do not require a Token, while verification and real booking, order, or payment actions do require a TourMind Skill Token.
For hotels, the workflow resolves a location, point of interest, or hotel with search_location, searches candidates with search_hotels, and retrieves live rates with batch_query_room_rates or query_room_rates. It can show property and room information with get_hotel_detail, then recheck the selected room's price, availability, and cancellation terms with check_room_availability. After the user provides the guest's legal name and contact email and confirms, it creates a booking with create_booking; query_booking checks it, while pay_order or an eligible cancel_booking action requires separate confirmation. For flights, it resolves airports with search_airports, searches one-way, round-trip, or multi-city itineraries with search_flights, and rechecks an offer with verify_offer. After traveler and document details are provided and the booking is confirmed, it calls create_booking; query_order checks the order, and the workflow can create a payment link with create_payment and look up payment status. The flight skill does not support cancellation, changes, refund initiation, ancillary purchases, or manual ticketing.
- A leisure traveler planning a Tokyo stay wants to compare live room rates, room types, breakfast, and cancellation terms before choosing.
- A business traveler needs to compare flight schedules, airports, connections, cabin, baggage allowance, and total fares.
- A trip planner wants to pair flights and hotels around dates and budget, with the two cost estimates shown separately.
- A traveler who has selected a hotel or flight needs to recheck current price, availability, and booking details before placing an order.
- Someone with an existing booking needs to check its status and, after reviewing the method and amount, continue to payment.
How do you install or deploy this agent?
Install both skills in a supported agent client. The README provides this command to install them to all detected clients; an environment with npx is needed.
How do you use this agent?
After installation, ask the agent to search for a hotel or flight. For example:
Find a hotel in Tokyo for two adults from December 9 to December 13, 2026. Compare live room options and show the total price, meals, and cancellation terms. Do not book yet.Hotel and flight search work without a Token. To verify offers or perform real booking, order, or payment actions, provide a TourMind Skill Token in a trusted private conversation. The two skills use the same Token.
What are this agent's strengths and limitations?
- The hotel workflow includes live rate comparison, room details, final price and availability checks, and booking management.
- The flight workflow searches one-way, round-trip, and multi-city trips and compares schedules, connections, baggage, and total prices.
- The conversation flow requires confirmation for bookings, payments, and eligible hotel cancellations; searches do not create orders.
- Offer verification and real booking, order, and payment actions require a TourMind Skill Token and depend on the TourMind API.
- The flight skill does not support cancellation, changes, refund initiation, ancillary purchases, or manual ticketing.
- Flight offers must be selected within 20 minutes of search; after that, a new live search requires approval.
- The README does not state service pricing, API limits, or availability commitments.
How does this agent compare with similar options?
The README presents this repository as the Skills option for both personal and business hotel and flight use. It also names Hotel Booking AI MCP for personal hotel use, TourMind Booking MCP for business hotel use, and Flight Booking AI MCP shared by personal and business users. Those MCP repositories are alternatives if an MCP package is needed instead of Skills.
Key facts side by side with the most closely related agents.
| Agent | Source review | Form / cost | Stars | Updated | Language | Full support on |
|---|---|---|---|---|---|---|
| TourMind Travel Booking Skills This agent | 56 · Major gaps | Agent plugin / skill | ★ 2.1k | 9d ago | Python | ChatGPT · Codex · Claude Code |
| lazyagent Coding-Agent Monitor | 75 · Good | CLIFree | ★ 193 | 2d ago | Go | Codex · Claude Code |
| Agent Desktop | 85 · Good | CLIFree | ★ 1.8k | today | Rust | — |
| GoLive Launch Assistant | 82 · Good | Agent plugin / skillFree | ★ 1.3k | 7d ago | TypeScript | Codex · Claude Code |
How does FollowAgents rate this agent?
Why each dimension lost points
The README says search does not create a booking and requires explicit confirmation before booking, eligible hotel cancellations, and payment, so confirmation and controls on external effects are reasonably handled (2). Token guidance covers private delivery, avoiding disclosure, mode 600 file permissions, and stopping authenticated operations after a 401; however, the token is stored locally and encryption, rotation, and protection of other traveler or payment data are not described, keeping sensitive-data handling at 2 and least privilege at 1. The README identifies TourMind API calls and discloses Stripe fees and that a payment link does not prove payment, but the full data flow and third-party processing scope are not documented (2). Confirmation is required for cancellation, but no recovery path is described (rollback 1). GitHub support and email contacts are listed, but publisher identity and accountable ownership are not verified (source attribution 1).
The hotel and flight flows separate search, verification, booking, and payment; the README also distinguishes candidate hotel prices from live rates and describes flight quote expiry and payment-state constraints (self-consistency 2). It gives a TourMind Token, an installation command, and some payment-channel constraints, but the provided material contains no runtime skill files or evidence of external-service availability (dependency availability 1). Validator test excerpts show some warnings and error conditions, but the implementation of failure handling and flight failure messages are not provided (failure messages 1).
The README covers consumer and business users, hotels and flights, combined itineraries, and natural-language requests, with concrete scenarios (audience and scenarios 3). It clearly states unsupported flight actions such as cancellation, changes, refunds, extras, and manual ticketing, and documents some workflow and payment constraints; runtime skill instructions are absent, limiting detail (capability boundaries 2). Examples and process descriptions specify when to search, verify, and obtain booking or payment confirmation (trigger precision 2). It lists supported agents, installation methods, and macOS/Linux token-file permissions, but runtime files needed to check environment fit are not supplied (environment fit 2).
The README is organized into capabilities, integrations, installation, examples, booking flows, security, FAQ, and support, though runtime skill instructions are missing (information architecture 2). It provides npx and agent-assisted installation plus token setup guidance, but no runtime files or platform-specific troubleshooting content are included (install notes 2). Tests require skill names to match directories and versions to stay synchronized, showing naming constraints; the actual frontmatter is not provided (naming stability 2). There are several hotel, flight, and combined-trip examples and an FAQ link, but not the FAQ contents (examples and FAQ 2). The README states unsupported flight operations and payment constraints (known limitations 2). The MIT license text is complete (license 3). A release badge and synchronized-version test are present, but no version history or changelog is included (versioning and changelog 1). GitHub issues and contact channels are listed, but maintenance ownership and update commitments are unclear, and publisher identity is unverified (maintenance responsibility 1).
Examples request useful details such as total prices, itineraries, baggage, cancellation terms, tradeoffs, and confirmation steps; actual output behavior and runtime instructions are not provided, so usability is 2. Live comparison, verification, booking, and hotel-management flows offer coherent travel-booking value supported by README detail (marginal value 2). Several authenticated operations require a TourMind Token, and Stripe has a disclosed non-refundable fee; there is no evidence about costs, coverage, or savings, so cost-benefit is 1.
The README gives specific claims about booking steps, quote timing, fees, limits, and token handling, and the test excerpts show some validator rules (claim traceability 2). Test files and a CI workflow indicate structural and validator checks exist, but they were not executed; README service claims also lack an independent second source in the supplied material (cross-source corroboration 1). The documentation distinguishes cached candidate prices from live bookable rates and payment links from completed payments, and lists limits, but runtime files are unavailable for verification (fact-inference separation 2).
- This review uses only the supplied README, license, CI configuration, and test excerpts. Runtime skill instructions and full implementation were not provided; tests were not run and service claims were not independently verified.
- Bookings, cancellations, and payments have external transaction effects. Before proceeding, check final prices, traveler details, cancellation terms, and payment method, and continue only after confirmation. The token is stored in a local file; keep it out of public conversations, logs, and shared materials.
Recent releases
Latest entries from the project's GitHub Releases. Notes are quoted from upstream as written, not rewritten or tested by FollowAgents.
-
v1.0.8 TourMind Booking Skills v1.0.8
Release notes excerpt
Highlights - Packages hotel and flight booking as two installable Skills in one repository: tourmind-booking and flight-booking-ai. - Adds live flight search, fare verification, booking, order lookup, and third-party payment workflows with confirmation and payment-safety controls. - Shares one TourMind Skill Token across both Skills, with current-directory priority and sibling-Skill fallback. - Unifies personal and business hotel flows, including business-permission handling, occupancy validation, batch rate checks, localized responses, and localized result-page URLs. …View on GitHub ↗ -
v1.0.4 TourMind Booking Skill v1.0.4
Release notes excerpt
Highlights - Removed the three embedded demo GIFs from the distributed source package. - Removed the Demo section from all four localized READMEs. - Reduced the current repository snapshot from about 11 MB to about 1.06 MiB. - Added a validation rule preventing demo GIFs from being redistributed. - Updated the declared Skill version to 1.0.4. What's Changed * docs: make skill installation client-neutral by @RemRam1 in https://github.com/tourmind-com/Tourmind-Booking-Skills/pull/2 …View on GitHub ↗ -
v1.0.0 TourMind Booking Skill v1.0.0
Release notes excerpt
TourMind Booking Skill v1.0.0 brings end-to-end hotel search and booking to AI agents through the TourMind API. Highlights - Resolve cities, hotels, landmarks, stations, addresses, ski areas, and other POIs. - Search up to 20 candidates, query live room products, and present the five best verified hotels. - Compare live rates across leading OTAs and hotel suppliers. - Recheck price and availability before booking. - Create bookings, manage orders and cancellations, and start Stripe, WeChat Pay, or Alipay payments. …View on GitHub ↗