executing-plans
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to "Do exactly what the step says" when processing plan tasks, creating a vulnerable loop for instructions embedded in data.
- Ingestion points: Task steps processed during the "Loop" defined in SKILL.md.
- Boundary markers: Absent. The skill provides no delimiters or instructions to ignore or validate embedded instructions within steps.
- Capability inventory: The skill facilitates general task execution based on the content of plan steps, providing a wide context for potential malicious instructions.
- Sanitization: Absent. There are no requirements for sanitizing or filtering step content before the agent executes it.
Audit Metadata