ClawAudit verdict

gan-evolution-engine

88
🟢 Trusted
Low risk — reviewed by ClawAudit, behavior matches stated purpose

A skill evolution framework using GAN-like LLM generation and fitness evaluation on local skill content; executionSinkDetected is from Python evolution script invocations operating on local data and the evomap credentials are for the skill submission platform integration.

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

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

25
security
70
transparency
70
maintenance

Findings (7)

Pattern match critical

Possible hardcoded credential

README.md · code · SECRET="your_node_secret

Pattern match medium

References agent configuration files

evolved_demo/self-improving-agent-v2/references/examples.md · code · CLAUDE.md

Pattern match medium

References agent memory files

evolved_demo/self-improving-agent-v2/references/openclaw-integration.md · code · MEMORY.md

Pattern match medium

pip3 install — installs Python packages at runtime

README.md · code · pip3 install

Pattern match medium

subprocess execution — runs system commands from Python

scripts/discriminator.py · prose · downgraded · subprocess.run(

Pattern match low

Python os.environ.get — reads environment variable

scripts/gan_evolution.py · prose · downgraded · os.environ.get(

Pattern match low

Python os.getenv — reads environment variable

scripts/generator.py · prose · downgraded · os.getenv(

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class B/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.

Is this flag fair?

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