postiz

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

• Third-party content exposure detected (high risk: 0.90). The skill instructs the agent to call integration tools like postiz integrations:trigger <integration-id> <method> (e.g., getFlairs for Reddit, getPlaylists for YouTube) to fetch dynamic, user-generated content from public third‑party platforms and then use that output (flair IDs, playlist IDs, etc.) directly in subsequent post-creation steps (see INTEGRATION_TOOLS_WORKFLOW.md and SKILL.md), which clearly exposes the agent to untrusted third-party content that can materially influence its actions.