Caveat verdict

codex-autopilot

🟢 Trusted

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

Multi-model coding automation orchestrator using tmux and launchd that installs from a public GitHub repo; tasks are routed to user-configured AI services and notifications go to user-owned Discord/Telegram channels.

⚠ 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.

security

transparency

maintenance

Permission integrity

Installs packages at runtime — transitive dependencies are not auditable

package_install

Findings (13)

Pattern match critical

Possible hardcoded credential

SKILL.md · code · token: "YOUR_BOT_TOKEN

Pattern match high

Pipe-to-shell pattern (curl | sh) — supply chain attack vector

code-review/SKILL.md · prose · downgraded · curl * \| bash

Pattern match high

Pipe-to-shell pattern (wget | sh)

code-review/SKILL.md · prose · downgraded · wget * \| sh

Pattern match high

Pipe to bash — executes piped content as shell commands

code-review/SKILL.md · prose · downgraded · | bash

Pattern match high

Pipe to sh — executes piped content as shell commands

code-review/SKILL.md · prose · downgraded · | sh

Pattern match high

Pipe to python — executes piped content as Python code

scripts/discord-notify.sh · prose · downgraded · | python3

Pattern match medium

subprocess execution — runs system commands from Python

lib/applescript_sender.py · prose · downgraded · subprocess.run(

Pattern match medium

References sudo — requests elevated privileges

code-review/SKILL.md · prose · downgraded · sudo

Pattern match medium

subprocess with shell=True — command injection vector

lib/done_checker.py · prose · downgraded · subprocess.run( cmd, shell=True

Pattern match low

Sets world-executable permissions

code-review/SKILL.md · prose · downgraded · chmod 777

Pattern match low

Popular HTTP library — network access

scripts/codex-status.sh · prose · downgraded · got

Pattern match low

Python os.environ.get — reads environment variable

scripts/test-agent.sh · prose · downgraded · os.environ.get(

Pattern match low

References agent configuration files

scripts/watchdog.sh · prose · downgraded · agent_config

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class A/B/D/F). 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