grilling
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides instructions for a structured interviewing process and does not contain any executable code, unauthorized network requests, or sensitive file access patterns.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input (goals and plans) which is used to drive tool execution for fact-finding, creating a potential attack surface.\n
- Ingestion points: User goals, plans, and decisions provided during the stress-testing rounds (SKILL.md).\n
- Boundary markers: Absent; the skill does not specify delimiters or instructions for the agent to ignore potentially malicious commands embedded in user-provided goals.\n
- Capability inventory: Environmental fact-finding involving filesystem and general tool access via sub-agents, along with command-like invocation of
/codebase-designand/domain-modeling.\n - Sanitization: Absent; the skill lacks specific validation or filtering steps for user-provided content before it influences tool usage.
Audit Metadata