JetBrains TeamCity EUVD-2026-102147

Exploitation requires network reachability to the TeamCity 'build messages protocol' - the channel used for communication involving build agents - and the ability to send a serialized message that bypasses the server's deserialization type whitelist (CWE-502). Affected deployments are TeamCity before 2026.2, 2026.1.4, and 2025.11.8; the fix is available from JetBrains. Limiting factors: in typical architectures the build messages protocol is an agent-facing channel, so real-world exploitation may depend on whether the endpoint is exposed to untrusted networks or whether the attacker can impersonate/reach an agent connection - the CVSS asserts PR:N (no authentication), but operators should verify exposure of this protocol port. CISA SSVC marks the attack as not automatable.

Signals are mixed and point to a serious-but-not-yet-actively-exploited flaw. The CVSS 3.1 vector (AV:N/AC:L/PR:N/UI:N, C:H/I:H/A:H, score 9.8) describes a network-reachable, unauthenticated, low-complexity RCE via unsafe deserialization (CWE-502), which is genuinely critical by design. However, the exploitation-intensity signals are low: EPSS is only 0.38% (31st percentile), the CVE is NOT in CISA KEV, no public POC is indicated, and CISA SSVC rates Exploitation as 'none' and Automatable as 'no' while confirming Technical Impact as 'total'. So the theoretical impact is maximal but observed/predicted exploitation is currently low - this is a patch-now priority for internet-exposed or CI/CD-critical TeamCity servers, but there is no evidence of in-the-wild attacks yet. One tension to verify: the CVSS claims PR:N (unauthenticated), yet the flaw is in the 'build messages protocol' between TeamCity server and build agents, which in many deployments is an authenticated/agent-trust channel - confirm with JetBrains whether a truly unauthenticated attacker can reach the vulnerable deserialization path, as the practical attack surface may be narrower than AV:N/PR:N implies. SSVC 'Automatable: no' is consistent with the latter interpretation.

An attacker who can reach the TeamCity build messages protocol endpoint sends a crafted, malicious serialized message that bypasses the type whitelist, triggering unsafe deserialization and arbitrary code execution on the TeamCity server. Because TeamCity orchestrates CI/CD, successful RCE yields control over build pipelines, secrets, and artifacts. No public exploit code is identified at time of analysis, and CVSS indicates low attack complexity with no user interaction.