walkthrough
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill contains instructions that direct the agent to ingest untrusted data from external sources to improve accuracy, which introduces a potential injection surface.
- Ingestion points: The instructions in SKILL.md (Step 1 and Gotchas) direct the agent to 'read the config, open the page, look at the code that renders the UI' and 'verify the actual menu names, flags, or field names' from the local machine or repository.
- Boundary markers: There are no instructions or boundary markers provided to ensure the agent ignores or treats content within these files as untrusted data.
- Capability inventory: The skill is designed to produce markdown instructions for the user, including copy-pasteable
bashcode blocks which are intended for human execution. - Sanitization: No sanitization, escaping, or validation steps are required by the skill before processing the external data sources.
Audit Metadata