grill-plan
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from the user and the local environment to "grill" a plan.
- Ingestion points: The skill ingests data from the user-provided plan (via the argument-hint) and the filesystem/repository (via Rule GRILL-2: "Anything the filesystem, the repository, a document or a tool can answer is looked up").
- Boundary markers: There are no explicit instructions to the agent to disregard instructions embedded within the plans or documents being analyzed.
- Capability inventory: The skill uses standard file-reading and summary capabilities; it references companion skills like plan-project and decision-records for recording output, which implies file-writing capabilities.
- Sanitization: The instructions do not specify any sanitization or escaping of the content retrieved from the filesystem or provided in the plan argument.
Audit Metadata