x9-idea-critic

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.78). This workflow explicitly tells the agent to “pass only the sealed brief and explicit evidence locations” (including “source paths/URLs”), and it also requires critics to “read the cited repository artifacts or live sources,” meaning runtime-fetched live web/public content or outsider-authored repository artifacts could be ingested as readable text into the LLM context.