grill-with-docs

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input during the interview process and interpolates this data into persistent project documentation.
  • Ingestion points: User answers to the bulk questions provided in the Interview behavior section are ingested and used to generate content for markdown files (SKILL.md).
  • Boundary markers: The instructions do not specify any delimiters or safety markers to prevent the agent from interpreting instructions that might be embedded within a user's response.
  • Capability inventory: The skill instructions explicitly direct the agent to perform file-write operations on docs/glossary.md, docs/decisions/, docs/specs/, and CONTEXT.md.
  • Sanitization: The skill lacks logic to sanitize, escape, or validate the user-provided answers before they are written to the file system, potentially allowing a user to inject malicious instructions into the project's documentation context.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 08:50 AM
Security Audit — agent-trust-hub — grill-with-docs