openserv-multi-agent-workflows

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly instructs research agents to fetch and ingest public, user-generated content from open web and social platforms (e.g., blog-pipeline and social-media-pipeline tasks: "Research the topic using X/Twitter" and "searches X/Twitter and web", plus youtube-to-blog which ingests YouTube URLs), so agents will read/interpret untrusted third-party content and could be exposed to indirect prompt injection.