ic-rooms

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow uses the room coordination log via ic_rooms_read/ic_rooms_send (MCP/A2A) to ingest other members’ committed turn text into the agent context, and those turns are outsider-authored free text (not authored by the operating user).

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). INSTALL.md contains explicit install commands (curl / Invoke-WebRequest) that download https://www.immersivecommons.com/skills/ic-rooms/SKILL.md into the agent's skills folder — this remote SKILL.md would be loaded as the agent's instruction content at runtime and thus directly controls prompts.