ClawAudit verdict
onescience-skill
onescience-skills
OneScience workflow orchestrator routing AI4S research tasks to defined sub-skills; executionSink reflects documented SLURM job submission matching the stated purpose.
⚠ 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.
Findings (5)
Uses eval() — can execute arbitrary code
onescience-debug/references/model_test.md · code · eval(
subprocess execution — runs system commands from Python
onescience-debug/references/e2e_pipeline_test.md · code · subprocess.run(
Python os.environ.get — reads environment variable
onescience-coder/assets/datapipes/tjweather.md · code · os.environ.get(
Popular HTTP library — network access
onescience-debug/references/model_test.md · code · got
Accesses .ssh directory
onescience-installer/SKILL.md · prose · downgraded · .ssh/
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?
Thanks — recorded.