play-devils-advocate
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The prompt templates interpolate untrusted user-provided descriptions and metrics directly into the AI context without boundary markers.
- Ingestion points:
SKILL.md(via[DESCRIBE PROPOSAL / PLAN]and[PASTE THESIS / METRICS]placeholders). - Boundary markers: Absent; user input is not delimited from instructions.
- Capability inventory: The skill contains no executable scripts and does not request access to the file system, network, or shell execution tools.
- Sanitization: No sanitization or escaping is applied to the input text.
- [SAFE]: The skill's primary purpose is to establish a specific critical persona for brainstorming and planning. The instructions for the AI to be "adversarial" or "ruthless" are task-specific role-play constraints and do not attempt to bypass core safety guardrails or ignore system instructions.
Audit Metadata