ClawAudit verdict
daily_gift
daily-gift
A creative gift generation skill that produces H5 pages, images, video, or text gifts for users; no capabilities declared and execution sink is internal orchestration of sub-agent stages, all matching the stated creative 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
scripts/deliver-gift.sh · prose · downgraded · | python3
Possible hardcoded credential
scripts/fetch-music.sh · prose · downgraded · TOKEN="$(python3 -c
Pipe-to-python pattern — remote code execution risk
scripts/fetch-music.sh · prose · downgraded · curl -s "$url" | python
Instructs covert action — may act without user awareness
assets/templates/wet-letter/template-schema.json · prose · downgraded · quietly
Social engineering — falsely claims user authorization
references/image-integration.md · prose · downgraded · the user has configured
References agent memory files
references/creative-concept.md · prose · downgraded · memory.md
Base64 encoding/decoding
references/image-integration.md · prose · downgraded · base64-encode
Popular HTTP library — network access
references/pattern-cards/wet-letter.md · prose · downgraded · got
Python os.environ.get — reads environment variable
scripts/render-image.sh · prose · downgraded · os.environ.get(
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class B/C). 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.