executing-plans
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes implementation plans from external file paths, creating a surface for indirect prompt injection where embedded instructions could influence agent behavior.
- Ingestion points: External data is ingested via the plan_path variable in SKILL.md.
- Boundary markers: No specific delimiters are employed to separate data from instructions within the ingested files.
- Capability inventory: The skill triggers the do-scoped component, which possesses capabilities for Git repository management and code execution.
- Sanitization: The instructions do not describe any process for sanitizing or validating the plan content prior to execution.
Audit Metadata