IntelliJ IDEA CVE-2026-75054

Exploitation requires the victim to open an untrusted project in IntelliJ IDEA - the crafted project must be loaded into the IDE, providing the 'User Interaction Required' (UI:R) condition. The OpenAPI preview proxy feature must process attacker-controlled server URL values from within the project's OpenAPI specification file. The attack vector is Local (AV:L): the attacker cannot exploit this remotely without first delivering a malicious project to the victim's filesystem. Limiting factors: the victim must actively open the project; the SSRF reaches only network resources the developer's machine can access (internal network, localhost, cloud metadata services - not arbitrary internet services from a server context); no persistence or code execution is implied by the CVSS impact metrics (C:H, I:N, A:N). Default IntelliJ IDEA installations with the OpenAPI preview proxy enabled are susceptible when untrusted projects are opened.

The CVSS 3.1 base score of 6.3 (Medium) is consistent with the vector: local delivery, no privileges required of the attacker, but mandatory user interaction and a scope change driving high confidentiality impact. The local attack vector (AV:L) is the primary risk-limiting factor - exploitation requires the victim to open a crafted project, making this a targeted social-engineering or supply-chain scenario rather than an opportunistic network-level attack. No EPSS data was supplied, and the absence of a KEV listing and public POC suggests exploitation has not been observed at scale. Real-world risk is meaningful for developer-targeted attacks: IDE users routinely open third-party repositories, and a malicious open-source project could silently trigger SSRF against the developer's internal network or cloud metadata service. No conflicting signals between the CVSS vector and the description were identified; the local, interaction-required model is consistent with the 'untrusted project' exploitation primitive described.

An attacker crafts a malicious IntelliJ IDEA project containing an OpenAPI specification with server URLs pointing to an internal network endpoint - such as a cloud provider metadata service (e.g., http://169.254.169.254/latest/meta-data/) or an internal CI/CD system. The victim clones or receives the project and opens it in IntelliJ IDEA, causing the OpenAPI preview proxy to issue an HTTP request to the attacker-controlled target. The response content, potentially including IAM credentials or internal service data, becomes accessible to the attacker. No public POC has been identified at time of analysis.