Network-AI (npm: network-ai) versions 5.12.2 through 5.13.3 fail to enforce the configured authorization check (checkAuth/secret) on the ApprovalInbox GET read routes. This allows unauthenticated actors to retrieve sensitive approval request information via the GET /approvals/?status=all, GET /approvals/:id, GET /approvals/stats, and GET /approvals/sse endpoints. The disclosed data includes full ApprovalEntry content such as shell-command strings, file paths, justifications, and risk levels. The responses also include a hardcoded Access-Control-Allow-Origin: * header, which enables cross-origin resource sharing from any website, increasing the risk of data leakage. This issue represents an incomplete fix of a previously known vulnerability (GHSA-mxjx-28vx-xjjj). No official patch or remediation level has been published yet.

Unauthenticated attackers can access sensitive approval request details, including potentially dangerous shell-command strings and file paths. The permissive CORS header allows any website to read this data if a user visits it, increasing the risk of data exfiltration and unauthorized disclosure of sensitive operational information. This could lead to further exploitation or operational compromise depending on the exposed data.

Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until an official fix is available, operators should consider restricting network access to the affected endpoints, disabling the ApprovalInbox GET routes if possible, or implementing external access controls to prevent unauthenticated access. Review and monitor for updates from Jovancoding regarding an official fix.