sup-writing-plans
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted software specifications to generate implementation plans, creating a surface for injection attacks.
- Ingestion points: The skill processes a "spec or requirements" document to define tasks (file:
SKILL.md) and a reviewer prompt template also processes external specifications (file:plan-document-reviewer-prompt.md). - Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish between the input specification and the skill's internal logic.
- Capability inventory: The skill allows writing files to the
.llm/plans/directory and facilitates shell command execution and code creation through its "Execution Handoff" feature (file:SKILL.md). - Sanitization: The skill lacks mechanisms to sanitize or validate the content of the provided specifications before they are incorporated into the plan.
Audit Metadata