grill-with-docs

Pass

Audited by Gen Agent Trust Hub on Jul 13, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill analyzes codebase content and existing documentation from a target repository to challenge designs and update glossaries. This represents an indirect prompt injection surface where instructions embedded in the codebase could influence agent behavior.
  • Ingestion points: Reads codebase and documentation files (such as CONTEXT.md and ADRs) in the target repository as specified in SKILL.md.
  • Boundary markers: The instructions do not define specific delimiters or instructions to ignore potential commands embedded in the codebase content.
  • Capability inventory: The skill is capable of writing documentation files (CONTEXT.md, docs/adr/) and performing git commits.
  • Sanitization: No specific sanitization or filtering of codebase content is mentioned before it is processed or used to drive documentation updates.
  • [COMMAND_EXECUTION]: The skill is instructed to commit documentation updates in the target repository, which involves executing shell commands (e.g., git commit) based on the decisions reached during the session.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 13, 2026, 09:53 AM
Security Audit — agent-trust-hub — grill-with-docs