CVE-2026-49377: JetBrains TeamCity Sensitive Data Exposure via Default Agent Parameters

Technical summary

CVE-2026-49377 is an information disclosure vulnerability in JetBrains TeamCity stemming from improper handling of default agent parameters. The flaw allows authenticated users to retrieve sensitive data that is inadequately protected in the agent configuration layer. The vulnerability is classified under CWE-526 (Exposure of Sensitive Information to an Unauthorized Actor) and carries a CVSS 3.1 score of 4.3 (Medium severity). The attack vector is network-based, requires low complexity, and necessitates prior authentication (PR:L), resulting in limited confidentiality impact with no integrity or availability consequences.

Business impact

For organizations running TeamCity as a continuous integration/continuous deployment platform, this vulnerability poses a data leakage risk to authenticated users—including potentially compromised internal accounts or malicious insiders. Exposed sensitive data from agent parameters could include API keys, credentials, environment configurations, or other CI/CD secrets that downstream systems depend on. While the direct blast radius is limited to information disclosure, exfiltrated credentials or configuration details could enable lateral movement or unauthorized access to dependent systems. The medium severity rating reflects that exploitation requires prior authentication, reducing but not eliminating organizational risk.

Affected systems

JetBrains TeamCity versions before 2025.11.2 are affected. All deployment architectures (cloud, on-premises, self-hosted) running vulnerable versions are in scope. The issue impacts the agent parameter subsystem specifically, meaning any TeamCity installation with active build agents is potentially affected. Users with authentication access—developers, CI/CD engineers, administrators, or compromised internal accounts—can trigger the exposure.

Exploitability

This vulnerability requires valid TeamCity login credentials to exploit; unauthenticated remote exploitation is not possible. The attack complexity is low, meaning an attacker with valid access can reliably retrieve sensitive data through normal agent configuration queries or API interactions without special conditions. There is no known public exploit or KEV (Known Exploited Vulnerability) designation. However, the ease of exploitation post-authentication makes this a concern for organizations with weak access controls or credential compromise scenarios.

Remediation

Upgrade to JetBrains TeamCity 2025.11.2 or later to patch the default agent parameter exposure. Organizations unable to patch immediately should restrict access to TeamCity through network segmentation, enforce multi-factor authentication for all user accounts, audit and rotate any credentials or API keys that may have been exposed via agent parameters, and monitor TeamCity logs for suspicious access to agent configuration endpoints.

Patch guidance

Apply the TeamCity 2025.11.2 release or later as soon as feasible. Consult the official JetBrains security advisory and release notes to confirm patch applicability to your specific deployment model (cloud vs. on-premises). Test patches in a non-production environment before rollout. If immediate patching is not viable, prioritize credential rotation and access restriction as interim mitigations.

Detection guidance

Monitor TeamCity audit logs for unusual access to agent configuration pages or APIs, particularly by non-administrative accounts or service accounts. Search logs for queries to agent parameter endpoints (e.g., GET requests to /app/rest/agents or related API paths). Correlate access patterns with known user roles to identify anomalous queries. If possible, enable verbose logging on the agent subsystem to capture parameter retrieval events. Additionally, review TeamCity's built-in audit trail for any access to sensitive configuration data by privileged accounts that may indicate account compromise.

Why prioritize this

Although rated Medium severity and requiring authentication, this vulnerability affects a critical infrastructure component (CI/CD platform) with high trust. Compromised agent parameters directly expose operational secrets. Organizations with strong authentication controls can deprioritize slightly; those with weaker credential hygiene or public-facing TeamCity instances should prioritize patching. The lack of KEV designation and low exploit complexity post-authentication suggests no active mass exploitation, but the sensitivity of CI/CD secrets warrants timely remediation.

Risk score, explained

The CVSS 3.1 score of 4.3 (Medium) reflects low-impact confidentiality loss (limited scope, no integrity/availability impact) tempered by the authentication requirement (PR:L). The network vector acknowledges remote access, and low complexity indicates ease of exploitation once authenticated. The score would be higher if the vulnerability enabled unauthenticated access or allowed credential modification; it remains medium because impact is strictly informational and access control gates exploitation.

Frequently asked questions