tg-hub

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly connects to Telegram and fetches user-generated messages (see SKILL.md and scripts/client.py: _fetch_history, sync/sync_all/refresh use telethon iter_messages to pull public/group/channel content) which is then read/searched/filtered by the agent (search/filter/today/recent APIs), so untrusted third‑party content can directly influence tool outputs and subsequent actions.