{"id":"CVE-2026-32625","aliases":["EUVD-2026-34046","GHSA-4pcc-j6m6-wcwx"],"status":"published","title":"LibreChat Exfiltrates Server Secrets via MCP Server URL Injection","description":"LibreChat is an enhanced ChatGPT clone that supports multiple AI providers. In versions up to and including 0.8.3, the Model Context Protocol (MCP) server integration resolves ${VAR} placeholders against the server's process.env during Zod schema validation of user-supplied MCP server URLs. Any authenticated user can create a malicious MCP server configuration with a URL pointing to an attacker-controlled domain containing environment variable references, causing the LibreChat server to connect to the attacker's server and transmit critical secrets such as CREDS_KEY, CREDS_IV, JWT_SECRET, and MONGO_URI in the request URL. This enables full compromise of the installation's cryptographic materials and database credentials without requiring administrative privileges. This is patched in version 0.8.4-rc1.","severity":{"version":"3.1","score":9.6,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:N","level":"critical","assessor":"cna","from":"cve"},"epss":{"day":"2026-11-11","score":0.02792,"percentile":0.86608,"model":"v2026.06.15"},"facts":{"in_kev":false,"known_exploited":false,"public_exploit":false,"fix_available":false},"cwes":["CWE-200"],"published":"2026-06-02T22:35:00Z","modified":"2026-10-28T02:30:46Z","products":[{"vendor":"danny-avila","product":"librechat"},{"vendor":"librechat","product":"librechat"}],"references":[{"url":"https://github.com/danny-avila/LibreChat/security/advisories/GHSA-4pcc-j6m6-wcwx","tags":["x_refsource_CONFIRM","exploit","Exploit","Mitigation","Vendor Advisory","advisory"]},{"url":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/32xxx/CVE-2026-32625.json","tags":["advisory"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-32625","tags":["advisory"]}],"sources":{"cve":[{"id":"CVE-2026-32625","title":"LibreChat Exfiltrates Server Secrets via MCP Server URL Injection","description":"LibreChat is an enhanced ChatGPT clone that supports multiple AI providers. In versions up to and including 0.8.3, the Model Context Protocol (MCP) server integration resolves ${VAR} placeholders against the server's process.env during Zod schema validation of user-supplied MCP server URLs. Any authenticated user can create a malicious MCP server configuration with a URL pointing to an attacker-controlled domain containing environment variable references, causing the LibreChat server to connect to the attacker's server and transmit critical secrets such as CREDS_KEY, CREDS_IV, JWT_SECRET, and MONGO_URI in the request URL. This enables full compromise of the installation's cryptographic materials and database credentials without requiring administrative privileges. This is patched in version 0.8.4-rc1.","status":"published","published":"2026-06-02T22:35:00.859Z","modified":"2026-06-03T14:07:46.872Z","cvss":[{"version":"3.1","score":9.6,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:N","level":"critical","assessor":"cna"}],"cwes":["CWE-200"],"updated_at":"2026-11-02T14:19:57.47301Z"}],"euvd":[{"id":"EUVD-2026-34046","description":"LibreChat is an enhanced ChatGPT clone that supports multiple AI providers. In versions up to and including 0.8.3, the Model Context Protocol (MCP) server integration resolves ${VAR} placeholders against the server's process.env during Zod schema validation of user-supplied MCP server URLs. Any authenticated user can create a malicious MCP server configuration with a URL pointing to an attacker-controlled domain containing environment variable references, causing the LibreChat server to connect to the attacker's server and transmit critical secrets such as CREDS_KEY, CREDS_IV, JWT_SECRET, and MONGO_URI in the request URL. This enables full compromise of the installation's cryptographic materials and database credentials without requiring administrative privileges. This is patched in version 0.8.4-rc1.","status":"published","published":"2026-06-02T22:35:00Z","modified":"2026-06-03T14:07:46Z","cvss":[{"version":"3.1","score":9.6,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:N","level":"critical"}],"updated_at":"2026-11-02T18:20:54.253343Z"}],"nvd":[{"id":"CVE-2026-32625","description":"LibreChat is an enhanced ChatGPT clone that supports multiple AI providers. In versions up to and including 0.8.3, the Model Context Protocol (MCP) server integration resolves ${VAR} placeholders against the server's process.env during Zod schema validation of user-supplied MCP server URLs. Any authenticated user can create a malicious MCP server configuration with a URL pointing to an attacker-controlled domain containing environment variable references, causing the LibreChat server to connect to the attacker's server and transmit critical secrets such as CREDS_KEY, CREDS_IV, JWT_SECRET, and MONGO_URI in the request URL. This enables full compromise of the installation's cryptographic materials and database credentials without requiring administrative privileges. This is patched in version 0.8.4-rc1.","status":"published","published":"2026-06-02T23:16:35.947Z","modified":"2026-07-22T19:10:00.12Z","cvss":[{"version":"3.1","score":9.6,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:N","level":"critical","assessor":"secondary"}],"cwes":["CWE-200"],"updated_at":"2026-11-02T14:12:22.022466Z"}],"osv":[{"id":"CVE-2026-32625","title":"LibreChat Exfiltrates Server Secrets via MCP Server URL Injection","description":"LibreChat is an enhanced ChatGPT clone that supports multiple AI providers. In versions up to and including 0.8.3, the Model Context Protocol (MCP) server integration resolves ${VAR} placeholders against the server's process.env during Zod schema validation of user-supplied MCP server URLs. Any authenticated user can create a malicious MCP server configuration with a URL pointing to an attacker-controlled domain containing environment variable references, causing the LibreChat server to connect to the attacker's server and transmit critical secrets such as CREDS_KEY, CREDS_IV, JWT_SECRET, and MONGO_URI in the request URL. This enables full compromise of the installation's cryptographic materials and database credentials without requiring administrative privileges. This is patched in version 0.8.4-rc1.","status":"published","published":"2026-06-02T22:35:00.859Z","modified":"2026-10-28T02:30:46.38844635Z","cvss":[{"version":"3.1","score":9.6,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:N","level":"critical"}],"cwes":["CWE-200"],"updated_at":"2026-11-02T15:45:01.548715Z"}]}}