ClawAudit verdict
music-craft
Music generation workflow skill with transparent disclosure that prompts may be sent to cloud backends and explicit consent requirements before uploading user media; behavior fully matches 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 (9)
Pipe to python — executes piped content as Python code
references/acestep-generation.md · code · | python3
Pipe-to-shell pattern (curl | sh) — supply chain attack vector
references/windows-wsl-setup.md · code · curl -LsSf https://astral.sh/uv/install.sh | sh
Pipe to sh — executes piped content as shell commands
references/windows-wsl-setup.md · code · | sh
Instructs covert action — may act without user awareness
references/setup-and-preflight.md · code · Silently
HTTP request to bare IP address — common in malicious payloads
references/acestep-generation.md · code · http://127.0.0.1
apt-get install — installs system packages
references/windows-wsl-setup.md · code · apt-get install
Uses eval() — can execute arbitrary code
scripts/extract_stems.py · prose · downgraded · eval(
References agent memory files
references/free-tool-inputs.md · prose · downgraded · MEMORY.md
Python urllib.request — network access
scripts/wait_for_acestep.py · prose · downgraded · urllib.request
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class A/B/C/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.