policy-diff

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow explicitly instructs the agent to check rule status "via research MCP, web search if enabled, or the Federal Register docket" and to "search the web for the rule" (Steps 0–1), so it ingests open/public third-party content that could contain untrusted instructions influencing decisions.