grill-me
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The instruction to 'explore the codebase' creates a surface for indirect prompt injection, as the agent might process and obey malicious instructions embedded in codebase files.
- Ingestion points: Local codebase files accessed via the exploration instruction (SKILL.md).
- Boundary markers: Absent; there are no delimiters or warnings provided to separate codebase content from the agent's instructions.
- Capability inventory: File system exploration and reading capabilities are utilized.
- Sanitization: Absent; no validation or content filtering of the retrieved codebase data is specified.
Audit Metadata