update-plan
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied context to update plan files, creating a surface where malicious instructions could potentially influence the agent.
- Ingestion points: New context is gathered from user arguments or the conversation history in SKILL.md.
- Boundary markers: The skill explicitly requires showing the user a delta and obtaining confirmation before any file writes in SKILL.md.
- Capability inventory: The skill is capable of reading and writing to PLAN.md and CONTEXT.md within the workspace.
- Sanitization: There are no explicit instructions for sanitizing or filtering the user-provided context before it is processed.
Audit Metadata