Ae1
High
- Category
- analysis-evasion
- Content
3. Merge the absorbed logic into the canonical `SKILL.md`.
- Confidence
- 100% confidence
- Finding
- Referenced artifact was not completely inspected
Security audit
Security checks for vulnerabilities and agentic risk
This skill does what it says, but it can delete a local skill folder and publish changes without requiring an explicit safety check first.
Install only if you are comfortable letting the agent modify local skill sources and publish updated skill behavior. Before using it, require the agent to show which skill will be kept and which folder will be retired, verify the merged SKILL.md, and archive or commit a backup before any deletion or ClawHub publish step.
3. Merge the absorbed logic into the canonical `SKILL.md`.
No suspicious patterns detected.