munger-perspective
Pass
Audited by Gen Agent Trust Hub on Nov 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The
SKILL.mdfile contains instructions for the agent to perform a version check by executing a shell command (git ls-remote). This command is used to query metadata from the official repository to alert the user of updates, representing a standard maintenance practice. - [EXTERNAL_DOWNLOADS]: The documentation references installation via
npxand points to the official GitHub repository for the skill. These resources are hosted on well-known platforms and are provided by the skill's author, fitting within the expected infrastructure of the extension. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted user queries to provide analysis. It implements boundary markers by clearly stating its persona is an AI-generated perspective and includes instructions to verify external facts through search tools, which mitigates potential risks associated with processing external data.
Audit Metadata