CVE-2025-52882 is a high-severity vulnerability affecting the Claude Code agentic coding tool extensions integrated into popular IDEs such as Visual Studio Code (VSCode) and its forks (Cursor, Windsurf, VSCodium), as well as JetBrains IDEs (IntelliJ, PyCharm, Android Studio). The root cause is a missing origin validation in the WebSocket implementation used by these extensions, classified under CWE-1385. This flaw allows attacker-controlled websites to establish unauthorized WebSocket connections to the vulnerable IDE extensions when a user visits a malicious webpage. In VSCode and its forks, exploitation enables an attacker to read arbitrary files accessible to the IDE, enumerate open files, capture selection and diagnostic events, and under limited conditions—specifically when a user has an open Jupyter Notebook and accepts a malicious prompt—execute arbitrary code. For JetBrains IDEs, the attacker can obtain selection events, lists of open files, and syntax error information, which could aid further targeted attacks or information leakage. The vulnerability affects Claude Code for VSCode versions 0.2.116 through 1.0.23 and JetBrains plugin beta versions 0.1.1 through 0.1.8. Claude released patches on June 13, 2025, addressing the issue by implementing proper origin validation to prevent unauthorized WebSocket connections. The vulnerability has a CVSS 4.0 base score of 8.8, reflecting its high impact on confidentiality and integrity, with network attack vector, low attack complexity, and requiring user interaction (visiting a malicious webpage and potentially accepting a prompt). No known exploits are currently reported in the wild. Users must update or uninstall vulnerable versions and restart their IDEs to mitigate the risk.

For European organizations, this vulnerability poses significant risks, especially for software development teams relying on the affected IDEs and Claude Code extensions. The ability for attackers to read arbitrary files and monitor IDE activities can lead to leakage of sensitive source code, intellectual property, and potentially credentials or configuration files stored or accessed within the IDE environment. The limited code execution capability in VSCode when Jupyter Notebooks are open further elevates the risk, potentially allowing attackers to execute malicious code within the developer's environment, leading to broader compromise. This can disrupt development workflows, cause data breaches, and facilitate supply chain attacks if compromised code is pushed to production. Given the widespread use of VSCode and JetBrains IDEs across European tech companies, research institutions, and government agencies, the vulnerability could impact a broad spectrum of sectors. The attack vector—via malicious webpages—means that phishing or drive-by attacks could be leveraged, increasing the threat surface. The absence of known exploits in the wild currently reduces immediate risk but does not eliminate it, as public disclosure and patch availability may prompt attackers to develop exploits. Organizations with remote or hybrid workforces may be particularly vulnerable due to increased exposure to web-based threats.

1. Immediate action should be to verify the versions of Claude Code extensions installed in all developer IDEs (VSCode, Cursor, Windsurf, VSCodium, IntelliJ, PyCharm, Android Studio). 2. Update the Claude Code for VSCode extension to version 1.0.24 or later and the JetBrains Claude Code [Beta] plugin to version 0.1.9 or later. If updates are not feasible immediately, uninstall the vulnerable extensions to eliminate exposure. 3. Restart the IDE after updating or uninstalling to ensure the patch takes effect. 4. Implement network-level controls to restrict outbound WebSocket connections from developer machines to untrusted domains, reducing the risk of unauthorized connections initiated by malicious webpages. 5. Educate developers about the risks of visiting untrusted websites while working in their IDEs, especially when using extensions that interact with external services. 6. Monitor IDE logs and network traffic for unusual WebSocket connections or file access patterns that could indicate exploitation attempts. 7. For organizations using Jupyter Notebooks within VSCode, enforce policies to avoid accepting prompts from untrusted sources and consider disabling notebook features if not required. 8. Incorporate vulnerability scanning and patch management processes specifically targeting development tools and their extensions to ensure timely updates in the future.