executing-plans
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to load and execute instructions from a user-provided or external plan file.
- Ingestion points: Step 1 instructs the agent to "Read plan file".
- Boundary markers: The skill lacks explicit boundary markers or delimiters (like XML tags or block markers) to separate the plan content from the system instructions.
- Capability inventory: The skill performs implementation tasks which typically include file system operations (TodoWrite) and execution of sub-skills for completing development branches.
- Sanitization: There is no explicit sanitization process mentioned, although the instructions include a "review critically" step to identify concerns.
Audit Metadata