recipe-plan-review

Pass

Audited by Gen Agent Trust Hub on Jul 11, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes internal management scripts using python3 .plan/execute-script.py. These scripts are used for configuration expansion and metadata persistence within the .plan directory. These are internal tools associated with the skill's infrastructure.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests content from archived request.md files and source code artifacts to perform analytical reviews. This constitutes an attack surface where instructions embedded in archived data could potentially influence the agent's logic.
  • Ingestion points: Reads .plan/local/archived-plans/{selected}/request.md and files listed in references.json.
  • Boundary markers: None identified; the skill processes raw file content during the comparison phase.
  • Capability inventory: The skill uses Read, Glob, Bash (for internal scripts), and Skill (to trigger standard fix workflows).
  • Sanitization: No explicit sanitization or filtering of the processed content is documented.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 11, 2026, 07:12 AM
Security Audit — agent-trust-hub — recipe-plan-review