cw-debate
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user content which acts as an entry point for indirect instructions.\n
- Ingestion points: The skill ingests drafts, briefs, outlines, and project instructions during the 'Load Context' stage (SKILL.md).\n
- Boundary markers: There are no instructions in the skill defining the use of delimiters or 'ignore' warnings for the ingested draft text.\n
- Capability inventory: The skill spawns multiple reviewer agents and conducts follow-up challenge rounds based on the draft's content, creating a multi-agent execution pipeline (SKILL.md).\n
- Sanitization: The skill lacks instructions for validating or sanitizing the input data before it is passed to the reviewer agents.
Audit Metadata