grilling
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to automatically perform environmental lookups based on the user's provided plan or decision.
- Ingestion points: User input describing plans, decisions, or ideas which the agent is instructed to 'grill' (
SKILL.md). - Boundary markers: There are no instructions providing delimiters or 'ignore' warnings for the data ingested from the user.
- Capability inventory: The skill explicitly authorizes the agent to 'dispatch a sub-agent' to query the environment, including the filesystem and other tools (
SKILL.md). - Sanitization: The instructions do not define any sanitization, filtering, or validation for user-provided data before it is used to trigger tool or filesystem operations.
Audit Metadata