military-flights

ReviewAudited by ClawScan on May 10, 2026.

Overview

This travel-search skill is not clearly malicious, but it should be reviewed because it can install an unpinned global CLI and persist raw travel queries locally.

Install only if you trust the flyai CLI package and are comfortable with a global npm install. Ask the agent to get approval before installing or rerunning searches, and disable or delete the local .flyai-execution-log.json file if you do not want travel or military-discount search details stored.

Findings (3)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

A global npm install can run third-party package code and change the user's system before any flight search is performed.

Why it was flagged

The skill tells the agent to install an unpinned global npm package at runtime instead of using a declared, reviewed install mechanism.

Skill content
If flyai-cli is not installed, install it first... npm i -g @fly-ai/flyai-cli
Recommendation

Require explicit user approval before installation, pin the package version, document the package provenance, and prefer a declared install spec or user-run setup.

What this means

The agent may rerun external flight searches instead of stopping promptly, causing delays or extra provider calls.

Why it was flagged

The skill makes repeated CLI execution the recovery path when expected booking links are absent, but it does not clearly define a retry limit.

Skill content
Self-test: If your response contains no `[Book](...)` links, you violated this skill. Stop and re-execute.
Recommendation

Add a small retry cap and ask the user before repeating failed or empty searches.

What this means

Travel routes, dates, and military/veteran-related search context could remain in a local file and be exposed through shared folders, backups, or later agent tasks.

Why it was flagged

The runbook records raw user queries and command history, then appends them to a persistent local file if filesystem writes are available.

Skill content
Agent maintains this log internally. Not shown to users... "user_query": "{raw input}" ... echo '{generation_log_json}' >> .flyai-execution-log.json
Recommendation

Make logging opt-in, redact sensitive query details, document the storage location and retention period, and provide a cleanup command.