setup
Fail
Audited by Gen Agent Trust Hub on Mar 21, 2026
Risk Level: HIGHCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill's primary function is to generate an
install-skills.shscript containing numerousnpx skills@latest addcommands. This pattern facilitates the downloading and immediate execution of code from external repositories. - [EXTERNAL_DOWNLOADS]: The skill references a massive catalog of external repositories. While some originate from trusted organizations (e.g., Anthropic, Microsoft, Google), a significant portion of the recommended code comes from unverified third-party accounts such as 'obra', 'supercent-io', 'inferen-sh', and 'mattpocock'. The skill encourages users to run the generated script immediately after creation.
- [COMMAND_EXECUTION]: In Step 2 of the skill workflow, the agent is instructed to execute bash commands (
ls ~/.claude/skills/andls .claude/skills/) to read the directory structure of the user's home directory and local project path.
Recommendations
- AI detected serious security threats
Audit Metadata