gpt-engineer

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

• Third-party content exposure detected (medium risk: 0.30). In scripts/run_codex_agent.py, the runner reads the outsider-provided --prompt-file contents or sys.stdin (user text) and concatenates it into delegated_prompt, which is then sent to the Codex delegate process via stdin (process.communicate(input=delegated_prompt, ...)).