executing-plans
Pass
Audited by Gen Agent Trust Hub on Jun 27, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill facilitates the ingestion and execution of instructions from an external plan file, creating a surface for indirect prompt injection if the plan originates from an untrusted source.
- Ingestion points: The workflow explicitly starts by instructing the agent to "Read plan file" (Step 1).
- Boundary markers: The skill includes behavioral mitigations, such as instructing the agent to "Review critically" and "identify any questions or concerns" before execution, and requiring human feedback between batches.
- Capability inventory: Depending on the contents of the plan, the agent may be directed to "Execute tasks" and "Run verifications," which typically involves file system modifications and command execution.
- Sanitization: No programmatic sanitization is defined; the skill relies on the agent's critical analysis and human-in-the-loop checkpoints to identify and block malicious instructions.
Audit Metadata