refine-plan
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process plan files (e.g.,
.turbo/plans/*.md). This creates an attack surface where instructions embedded within the plan data could attempt to influence the agent's behavior during the refinement loop. - Ingestion points: Step 1 resolves and reads plan files from the file system.
- Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the content of the plan files.
- Capability inventory: The skill can write to files via
/apply-findingsand execute 'instrumented experiments'. - Sanitization: No explicit sanitization or validation of the plan file content is mentioned.
- [DYNAMIC_EXECUTION]: The skill instructs the agent to 'run the instrumented experiment' to verify technical behaviors of dependencies or platforms in a temporary directory. This involves the dynamic generation and execution of code based on findings during the plan review process.
Audit Metadata