AI agent vulnerability · mcp-remote
CVE-2025-6514: mcp-remote exposed to OS command injection via untrusted MCP server connections
criticalCVSS 3.1 9.6GHSA-6xpm-ggf7-wc3p
mcp-remote is exposed to OS command injection when connecting to untrusted MCP servers due to crafted input from the authorizationendpoint response URL
Published Jul 9, 2025
Affected packages
mcp-remote
- Ecosystem
- npm
- Affected
- >= 0.0.5, < 0.1.16
- Fixed in
- 0.1.16
- Check
npm ls mcp-remote
Advisory details
From the GitHub Advisory Database (CC BY 4.0).
mcp-remote is exposed to OS command injection when connecting to untrusted MCP servers due to crafted input from the authorization_endpoint response URL
References
- https://nvd.nist.gov/vuln/detail/CVE-2025-6514
- https://github.com/geelen/mcp-remote/commit/607b226a356cb61a239ffaba2fb3db1c9dea4bac
- https://jfrog.com/blog/2025-6514-critical-mcp-remote-rce-vulnerability
- https://research.jfrog.com/vulnerabilities/mcp-remote-command-injection-rce-jfsa-2025-001290844
- https://github.com/advisories/GHSA-6xpm-ggf7-wc3p
How this page is made: the data panel comes from the GitHub Advisory Database, CISA’s KEV catalog, the NVD and FIRST EPSS, and updates twice a day. No analysis has been written for this entry yet. Methodology