grill-me
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill directs the agent to explore the codebase to answer questions, which introduces a surface for processing untrusted content.\n
- Ingestion points: Codebase files and user-provided design plans (SKILL.md).\n
- Boundary markers: Absent; the instructions do not define delimiters or specific warnings to ignore instructions found within project files.\n
- Capability inventory: Involves codebase exploration (file reading) and conversational interaction. No risky capabilities like file-writing or network access are explicitly used by the skill.\n
- Sanitization: Absent; the skill does not include logic to sanitize or escape data retrieved from the codebase.
Audit Metadata