grill-with-docs

Pass

Audited by Gen Agent Trust Hub on Aug 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to explore the project codebase and update persistent documentation (such as CONTEXT.md or Architecture Decision Records) based on its findings. This creates a surface where malicious instructions embedded in project files (e.g., in code comments or data files) could be ingested by the agent and influence its decision-making or the content it writes to persistent storage.
  • Ingestion points: Project source code, directory structures, and existing documentation files (referenced in SKILL.md under Execution Rules and Example Invocation Flow).
  • Boundary markers: Absent. The instructions provide no delimiters or specific directives to treat codebase content as untrusted data or to ignore instructions found within those files.
  • Capability inventory: The skill uses file system read access for exploration and file system write access to create or update files in the project root, docs/decisions/, or adr/ directories.
  • Sanitization: Absent. There are no explicit instructions for the agent to sanitize, validate, or filter the content retrieved from the codebase before interpolating it into the final documentation artifacts.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 17, 2026, 11:52 AM
Security Audit — agent-trust-hub — grill-with-docs