Caveat verdict

code-refactor-for-reproducibility

code-refactor-for-reproducibility-1

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

Research code refactoring skill that runs a local Python script with standard data science dependencies (numpy, pandas, scipy, pytest); no network access or exfiltration behavior present.

⚠ 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.

25
security
60
transparency
70
maintenance

Permission integrity

Performs file operations but does not declare file-accessing binaries

file_read

Installs packages at runtime — transitive dependencies are not auditable

package_install

Findings (5)

Pattern match high

Recursive delete from root or home — destructive command

scripts/main.py · prose · downgraded · rm -rf /

Pattern match high

Pipe to python — executes piped content as Python code

scripts/main.py · prose · downgraded · | python

Pattern match medium

subprocess execution — runs system commands from Python

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

Pattern match medium

apt-get install — installs system packages

scripts/main.py · prose · downgraded · apt-get install

Pattern match low

Python pathlib file read

SKILL.md · code · pathlib.Path("analysis.py").read_text(

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.

file_readpackage_install

Is this flag fair?

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