Caveat verdict

ui-agent

88
🟢 Trusted
No high-risk patterns surfaced by the deep scan — automated capability review, not behavioral proof.

Browser and desktop UI automation framework using Chrome DevTools Protocol and native OS accessibility APIs; capabilities are appropriate for the stated RPA and testing purpose with no exfiltration.

⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.

Automated static analysis — not a human review. Caveat flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.

0
security
90
transparency
90
maintenance

Permission integrity

Installs packages at runtime — transitive dependencies are not auditable

package_install

Findings (9)

Pattern match high

References sudo — requests elevated privileges

docs/SKILL.md · code · sudo

Pattern match high

apt-get install — installs system packages

docs/SKILL.md · code · apt-get install

Pattern match high

subprocess execution — runs system commands from Python

docs/SKILL.md · code · subprocess.run(

Pattern match high

Possible hardcoded credential

tests/final_test_e1_fixed.py · prose · downgraded · password='{password}

Pattern match medium

Opens WebSocket connection

SKILL.md · frontmatter · websocket

Pattern match medium

Popular HTTP library — network access

SKILL.md · code · got

Pattern match medium

os.system/popen — direct OS command execution

src/cdp_typer.py · prose · downgraded · os.system(

Pattern match medium

subprocess with shell=True — command injection vector

tests/final_test_t1.py · prose · downgraded · subprocess.run( f"echo 'UIAgent terminal final test' > {OUTPUT}", shell=

Pattern match low

Python os.environ.get — reads environment variable

src/chrome_session_vbox_fixed.py · prose · downgraded · os.environ.get(

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class D). Final tier capped at Caution — cannot be lifted by any downgrade, example-payload opt-in, or allowlist.

Permissions & capabilities

No declared permissions — minimal attack surface.

package_install

Is this flag fair?

Check another skill Browse the registry Auditing your own skills or configs? Use the API