test-martool-cli

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

• Third-party content exposure detected (high risk: 0.75). In Coolify mode, the agent reads OUTSIDER-authored free text via SSH from the remote container by executing node dist/cli/main.js usage --json and version --json, then parses result.stdout as JSON and injects it into the agent’s LLM context through the combined combined.json/stdout evidence path.