thread-management

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

• Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md shows commands that read and search chat threads (e.g., alma thread messages <thread-id> and the threads search API plus switching Telegram/Discord chats), which ingests user-generated, potentially untrusted third-party content that the agent would read and could influence actions.