ai-consultant

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

• Third-party content exposure detected (high risk: 0.80). The skill's required Phase 1 "Pre-Engagement Research" explicitly directs the agent to web search and ingest public third-party sources (e.g., press releases, job postings, engineering blogs, public repos) to build a company-briefing that the agent must use to drive assessment and recommendations, which clearly exposes it to untrusted user-generated/open-web content that can materially influence subsequent actions.