atk-cfl

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

• Third-party content exposure detected (high risk: 0.75). The required workflow is to run atk-cfl page view / atk-cfl search and feed the resulting Confluence page content (outsider-authored text from other users) into the agent’s LLM context via the tool’s stdout output (e.g., --content-only / -o plain).