ocs

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

• Third-party content exposure detected (medium risk: 0.30). The workflow monitors local cross-session messages and inboxes via ocs inbox and ocs read, which ingest outsider-authored messages from other local AI coding agents and channels.

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

• Potentially malicious external URL detected (high risk: 0.80). The installation instruction uses curl to fetch and execute a remote shell script from GitHub at runtime, meeting the threshold for remote code execution during setup/installation.