agora

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text can be ingested into the agent’s LLM context when the agent runs agora.py read / inbox / watch, because those commands load and render Markdown message bodies from ~/.agora/<repo-id>/threads/<thread>/*.md and ~/.agora/<repo-id>/inbox/<handle>/*.txt, which may have been posted by other agents/users.