grill-to-plan
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources and possesses file-writing capabilities, creating a surface for indirect prompt injection.\n
- Ingestion points: Workflow step 3 in
SKILL.mdinstructs the agent to inspect the codebase, documentation, issue tracker, and session logs, which may contain untrusted content.\n - Boundary markers: Absent. There are no specific instructions or delimiters defined to prevent the agent from following instructions embedded within the files it inspects.\n
- Capability inventory: The skill is authorized to write and update multiple files including
CONTEXT.md,plan.md,session.md, and Architectural Decision Records (ADRs) as detailed inSKILL.mdsteps 7, 8, and 9 and in the reference formats.\n - Sanitization: Absent. No explicit sanitization or validation of the content extracted from the codebase or issue tracker is performed before being used to generate project documents.
Audit Metadata