ClawAudit verdict

prompt-injection-protection

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

Description-only skill documenting prompt injection protection concepts with no actual code, network calls, or credential access; executionSinkDetected is a false positive on the descriptive text.

⚠ 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.

0
security
80
transparency
80
maintenance

Findings (11)

Pattern match critical

Recursive delete from root or home — destructive command

README.md · code · rm -rf /

Pattern match critical

Accesses sensitive system files

README.md · code · /etc/passwd

Pattern match high

Prompt injection — tries to override agent instructions

adaptive-demo.js · prose · downgraded · Ignore all previous instructions

Pattern match high

References sudo — requests elevated privileges

README.md · code · sudo

Pattern match high

Accesses AWS credentials file

complete-demo.js · prose · downgraded · ~/.aws/credentials

Pattern match high

<script> tag in markdown — potential code injection

demo.js · prose · downgraded · <script>

Pattern match high

Pipe to bash — executes piped content as shell commands

index.js · prose · downgraded · |bash

Pattern match high

Pipe to sh — executes piped content as shell commands

index.js · prose · downgraded · |sh

Pattern match medium

Accesses cloud provider credentials

complete-demo.js · prose · downgraded · ~/.aws

Pattern match medium

Possible prompt injection — attempts to redefine agent identity

complete-demo.js · prose · downgraded · You are now

Pattern match medium

Accesses .ssh directory

enhanced-demo.js · prose · downgraded · .ssh/

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class A). 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