The source exposes agent messaging and tool-permission capabilities over an all-interface HTTP listener, with permissive CORS and no authentication check in the inspected routes. This is a real unresolved exposure risk, but the inspected behavior does not establish malicious activity.
Static reason
No blocking static signals were detected.
Trigger
A user starts the server directly or invokes the daemon entrypoint.
Impact
A network-reachable caller may invoke agent behavior and grant tool permissions. The inspected source does not establish arbitrary code execution or malicious intent.
Mechanism
HTTP requests reach the message route, which calls the orchestrator, and the permission route can grant a named tool; the inspected routes show no authentication check.
Evidence
dist/index.jsdist/routes/messages.jsdist/routes/permissions.jsdist/middleware/cors.jsbin/cortex-daemon.js