Caveat verdict
fr24-ai
International flight booking skill for Flightroutes24 that calls documented API endpoints using configured credentials; the skill explicitly instructs the agent not to reveal API keys or expose them in conversation, consistent with a legitimate booking platform integration.
β Flagged for review β coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.
Automated static analysis β not a human review. Caveat flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.
Findings (7)
Possible hardcoded credential
references/setup-maintainer.md Β· code Β· SECRET='δ½ ηSHA512ηΎεε―ι₯
Pipe to python β executes piped content as Python code
README.en.md Β· prose Β· downgraded Β· | Python
Dynamic __import__('os') β Python OS command execution
scripts/config_keys.py Β· prose Β· downgraded Β· __import__("os")
subprocess execution β runs system commands from Python
scripts/booking_flow_test.py Β· prose Β· downgraded Β· subprocess.run(
Python os.environ.get β reads environment variable
config.py Β· prose Β· downgraded Β· os.environ.get(
Python urllib.request β network access
scripts/newapi_client.py Β· prose Β· downgraded Β· urllib.request
Popular HTTP library β network access
scripts/nl_scenario_test.py Β· prose Β· downgraded Β· got
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class D). Final tier capped at Caution β cannot be lifted by any downgrade, example-payload opt-in, or allowlist.
Permissions & capabilities
No declared permissions β minimal attack surface.
Is this flag fair?
Thanks β recorded.