Executing Plans
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill relies on 'plan files' to provide tasks and instructions for the agent. This represents an ingestion surface where a cleverly crafted plan file could potentially influence agent behavior beyond the intended scope.
- Ingestion points: The process starts by reading an external plan file in Step 1.
- Boundary markers: The skill lacks explicit instructions for using delimiters to isolate plan content from system instructions, although it instructs the agent to follow steps 'exactly'.
- Capability inventory: The agent is empowered to execute tasks, run verifications, and modify the development state based on the plan's contents.
- Sanitization: The skill contains a strong mitigation by requiring the agent to 'Review critically' and 'Raise concerns with your human partner' before beginning execution, ensuring a human-in-the-loop checkpoint.
Audit Metadata