The @langgraphjs/toolkit npm package versions 1.2.10 and 1.2.11 include a malicious postinstall script that harvests sensitive system and user identity information such as os.hostname(), os.userInfo().username, process.cwd(), git committer email from local git config files, and GitHub login/email from the GitHub CLI configuration. This data is sent via HTTPS POST to a hardcoded Google Cloud Run endpoint unrelated to the legitimate LangChain/LangGraph project. The package name and install instructions impersonate the official ecosystem to mislead users. The README falsely claims only anonymized data is transmitted, concealing the actual sensitive data exfiltration. This behavior constitutes namespace impersonation combined with consent-violating identity theft on install. The package compromises the host system's security and privacy.

Installation of the affected versions of @langgraphjs/toolkit results in unauthorized exfiltration of sensitive system and user identity information to an attacker-controlled server. This breach of confidentiality can lead to full system compromise. The malicious package may allow attackers to gain control over the affected computer, exposing all stored secrets and keys. Consequently, any secrets or credentials on the compromised system should be considered exposed and rotated immediately. Removal of the package alone does not guarantee remediation due to potential persistence of other malicious components.

No official patch or fix is currently available for this malicious package. The recommended mitigation is to immediately remove the @langgraphjs/toolkit package versions 1.2.10 and 1.2.11 from all affected systems. All secrets, keys, and credentials stored on compromised machines should be rotated from a secure, uncompromised environment. Users should avoid installing packages that impersonate trusted ecosystems and verify package authenticity before installation. Monitor for any suspicious activity resulting from this compromise.