atelier-grill-me

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPERSISTENCE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the repository's codebase to inform its interview logic and recommendations.
  • Ingestion points: The agent is instructed to read the files, the config, and existing patterns within the working directory.
  • Boundary markers: The instructions do not specify any delimiters or safety warnings to distinguish between file content and instructions.
  • Capability inventory: The skill has the capability to write decision records to the docs directory and update local journal files based on the interview outcomes.
  • Sanitization: There is no mention of filtering or sanitizing the content read from the codebase before it is used to generate recommendations.
  • [PERSISTENCE]: The skill attempts to maintain context across sessions by instructing the agent to append durable choices as decision entries in a local lessons journal file.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 08:48 AM
Security Audit — agent-trust-hub — atelier-grill-me