ClawAudit verdict
juicebox-v5
juicy
Comprehensive Juicebox V5 DeFi protocol skill collection covering contract deployment, queries, and UI generation; no evidence of exfiltration or malicious behavior in the content.
⚠ 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 (7)
Possible hardcoded credential
jb-deploy-ui/SKILL.md · code · token: '0x000000000000000000000000000000000000EEEe
<script> tag in markdown — potential code injection
jb-deploy-ui/SKILL.md · code · <script
Instructs covert action — may act without user awareness
jb-bendystraw/SKILL.md · code · silently
Data URI with base64 payload — may embed malicious content
jb-patterns/SKILL.md · code · data:application/json;base64,
Accesses sensitive environment variables
jb-bendystraw/SKILL.md · code · process.env.BENDYSTRAW_API_KEY
Base64 encoding/decoding
jb-patterns/SKILL.md · code · base64-encode
Makes HTTP request to external URL
jb-bendystraw/SKILL.md · code · fetch(`https://
Permissions & capabilities
No declared permissions — minimal attack surface.
Is this flag fair?
Thanks — recorded.