grill-with-docs
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes project-specific data which creates a potential surface for indirect prompt injection attacks. 1. Ingestion points: The skill is instructed in SKILL.md to trace request paths through source code and grep for reads in fields, types, and endpoints. It also reads existing CONTEXT.md and ADR files. 2. Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following instructions that might be embedded in the code or documentation it analyzes. 3. Capability inventory: The skill uses file system reading (grep) and file system writing/modification (updating CONTEXT.md and ADRs). 4. Sanitization: The skill does not implement validation or sanitization of the content extracted from the repository files before the agent processes it.
Audit Metadata