ZeroDay Findings
Sign inSign up

Findings

CVE-2026-42271

BerriAI LiteLLM Command Injection Vulnerability

Published May 7, 2026 · updated Jul 14, 2026 · BerriAI LiteLLM

LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.74.2 to before version 1.83.7, two endpoints used to preview an MCP server before saving it — POST /mcp-rest/test/connection and POST /mcp-rest/test/tools/list — accepted a full server configuration in the request body, including the command, args, and env fields used by the stdio transport. When called with a stdio configuration, the endpoints attempted to connect, which spawned the supplied command as a subprocess on the proxy host with the privileges of the proxy process. The endpoints were gated only by a valid proxy API key, with no role check. Any authenticated user — including holders of low-privilege internal-user keys — could therefore run arbitrary commands on the host. This issue has been patched in version 1.83.7.

Timeline

Detection, updates, and resolution for this record.

  1. May 7, 2026 · Catalog

    Detected

    LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.74.2 to before version 1.83.7, two endpoints used to preview an MCP server before saving it — POST /mcp-rest/test/connection and POST /mcp-rest/test/tools/list — accepted a full server configuration in the request body, including the command, args, and env fields used by the stdio transport. When c

  2. Jun 8, 2026 · CISA KEV

    Known exploited

    Apply mitigations per vendor instructions, follow applicable BOD 22-01 guidance for cloud services, or discontinue use of the product if mitigations are unavailable.

  3. Jun 22, 2026 · CISA KEV

    Remediation due

    Apply mitigations per vendor instructions, follow applicable BOD 22-01 guidance for cloud services, or discontinue use of the product if mitigations are unavailable.

  4. Jul 14, 2026 · Catalog

    Updated

    LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.74.2 to before version 1.83.7, two endpoints used to preview an MCP server before saving it — POST /mcp-rest/test/connection and POST /mcp-rest/test/tools/list — accepted a full server configuration in the request body, including the command, args, and env fields used by the stdio transport. When c

Required action

Apply mitigations per vendor instructions, follow applicable BOD 22-01 guidance for cloud services, or discontinue use of the product if mitigations are unavailable.

CISA due date Jun 22, 2026

Affected

  • cpe litellm/litellmversion unspecified
  • cpe redhat/openshift aiversion unspecified
  • cpe redhat/openshift ai3.4
  • vendor BerriAI LiteLLMversion unspecified

Sources

  • CISA KEVCVE-2026-42271

    BerriAI LiteLLM contains a command injection vulnerability that could allow any authenticated user, including holders of low-privilege internal-user keys, to run arbitrary commands on the host.

    cisa.gov
  • CVE.orgCVE-2026-42271

    Official CVE identifier

    cve.org
  • FIRST EPSSCVE-2026-42271

    30-day exploitation probability 92.6%

    first.org
  • Metasploit Modulesexploit_multi/http/litellm_mcp_test_cmd_injection

    Public module reference (excellent). Module source is not copied into this catalog.

    rapid7.com
  • NIST NVDCVE-2026-42271

    LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.74.2 to before version 1.83.7, two endpoints used to preview an MCP server before saving it — POST /mcp-rest/test/connection and POST /mcp-rest/test/tools/list — accepted a full s

    nvd.nist.gov

References