OpenCATS CVE-2026-49490

Attacker must possess valid authenticated credentials to the OpenCATS web application (CVSS PR:L) and be able to reach the Candidates DataGrid filtering endpoint over the network; the specific trigger is a crafted DataGrid filter request that targets the Tags column, which is normally marked non-filterable in the UI but is not enforced server-side. Exploitation requires OpenCATS at version 0.9.1a or later and does not require user interaction, administrative privileges, or any non-default configuration. The main limiting factor is the authentication requirement - internet-exposed OpenCATS instances without registration controls effectively reduce this to a near-unauthenticated condition, while instances limited to internal SSO meaningfully constrain the attacker pool.

The CVSS 4.0 vector (AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:N) reflects a network-reachable, low-complexity attack requiring low privileges and no user interaction, with high confidentiality and integrity impact and no availability impact - consistent with a classic authenticated SQLi reaching sensitive recruitment data. EPSS data was not provided, and the CVE is not in CISA KEV, so there is no evidence of mass or targeted exploitation yet. Authentication is required (PR:L), which lowers real-world risk versus an unauthenticated SQLi, but OpenCATS deployments commonly grant low-privilege recruiter or interviewer accounts to many internal users, so the bar to obtain qualifying credentials is realistically low. Overall this is a credible priority for any organization running OpenCATS 0.9.1a or later, even without exploitation telemetry.

An attacker with any valid low-privilege OpenCATS account (for example, a basic recruiter or interviewer login) navigates to the Candidates DataGrid and intercepts the filter request, then modifies it to apply a filter against the non-filterable Tags column containing a SQL payload such as a UNION SELECT. The server-side handler ignores the UI's filterable restriction, concatenates the payload into the underlying query, and returns or modifies arbitrary database contents - exposing candidate PII, resumes, and credentials stored in the ATS. No public exploit identified at time of analysis, but the path is straightforward given the documented bypass.