executing-plans
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes implementation plans from external files (Step 1), which creates a surface for instructions embedded in those files to influence agent behavior.
- Ingestion points: The skill reads external plan files during the 'Load and Review Plan' phase in
SKILL.md. - Boundary markers: The skill lacks explicit technical delimiters or boundary markers to isolate plan content from agent instructions.
- Capability inventory: The skill is authorized to execute all tasks defined in a plan, which typically includes file modifications, tool usage, and terminal command execution.
- Sanitization: The skill relies on 'critical review' by the AI and manual confirmation from the user (Step 1.3) to identify malicious or incorrect plans rather than technical sanitization or input validation.
Audit Metadata