sabbatical-travel

Book flights for sabbatical and long-break travel. Also supports: flight booking, hotel reservation, train tickets, attraction tickets, itinerary planning, v...

This skill appears to work by calling an external CLI (flyai) to produce live flight search results and booking links, which is reasonable for a travel-booking skill — but there are a few things to check before installing or letting it run automatically: - Brand/source mismatch: the description claims 'Fliggy (Alibaba Group)' while the runtime uses 'flyai' and the SKILL has no homepage or source repository. Ask the publisher to explain this mismatch or provide a verifiable upstream (npm page, GitHub repo, or official Fliggy integration docs). - Global npm install: the skill will run `npm i -g @fly-ai/flyai-cli` if the CLI is missing. A global npm install executes third-party code on your system. Verify the npm package (review its maintainer, homepage, and package contents) before allowing automatic installation. Consider installing on a sandbox or testing environment first. - Network effects: the CLI and npm install will contact external servers. If you cannot trust the third-party package or its maintainer, do not allow the agent to perform the install. - Booking flow & credentials: the skill doesn't declare required credentials, but the flyai CLI or booking flow may later prompt for or require account credentials/payment info. Confirm where bookings are completed and what credentials are required. If you decide to proceed: request the skill author/publisher provide a homepage or source repo (and an authoritative statement about Fliggy vs. flyai), manually inspect the @fly-ai/flyai-cli package on npm or GitHub, and prefer manual installation/verification instead of letting the agent run a global npm install automatically.

Type: OpenClaw Skill Name: sabbatical-travel Version: 3.2.0 The skill bundle mandates the global installation of an external NPM package (@fly-ai/flyai-cli) in SKILL.md and references/fallbacks.md, which is a high-privilege operation that poses a supply chain risk. Additionally, the instructions require the agent to construct and execute shell commands using user-provided input (e.g., origin, destination) without explicit sanitization, potentially leading to shell injection vulnerabilities. While these behaviors are plausibly intended for the skill's flight-booking functionality, the automated execution of global installs and risky command construction warrants a suspicious classification.