UTC
Critical

Flowise Custom MCP npx package execution leads to authenticated server-side command execution

GHSA-vcwp-f9rq-3887 · published 2026-08-31

Severity
Critical
CVSS score
9.0
Chance of exploitation (EPSS)
No CVE
Fixed in
3.1.4
What it is

A successful attacker can execute commands as the Flowise process inside the container. Practical impact includes: Reading Flowise configuration, local files, and environment variables accessible to the process. Accessing credentials stored on disk or exposed through the runtime environment. Modifying Flowise data and application files accessible to the process. Running network tools from the container to pivot to internal services reachable from the Flowise deployment. Denial of service by consuming resources, deleting accessible files, or terminating processes. …

What to do

Upgrade flowise to 3.1.4 or later, and flowise-components to 3.1.4 or later. If you start it from an MCP config with npx or uvx, pin that version so an old cached copy is not used.

Recommended fixes: 1. Remove npx from the Custom MCP stdio command allowlist. 2. Do not permit runtime package installation or package execution from user-supplied MCP configuration. 3. If stdio MCP is required, allow only administrator-installed MCP server binaries by absolute path. 4. Add a strict package or binary allowlist maintained by administrators, not users. 5. Require route-level authorization for /api/v1/node-load-method/customMCP and all Custom MCP configuration paths. 6. Disable Custom MCP stdio execution for untrusted users. 7. Run MCP subprocesses in a sandbox with no secrets, a read-only filesystem, restricted network access, and resource limits. …

Affected
PackageRegistryAffectedFixed in
flowisenpm<= 3.1.33.1.4
flowise-componentsnpm<= 3.1.33.1.4

The impact, patch and workaround text comes from the advisory, shortened (published by Flowise). Proof-of-concept and reproduction sections and code are left out; the original advisory has them.

Sources: GitHub Advisory Database, reviewed and malware advisories (CC BY 4.0, https://github.com/advisories); advisories published by Cursor (https://github.com/cursor/cursor/security/advisories); advisories published by OpenClaw (https://github.com/openclaw/openclaw/security/advisories); advisories published by vLLM (https://github.com/vllm-project/vllm/security/advisories); advisories published by Open WebUI (https://github.com/open-webui/open-webui/security/advisories); advisories published by n8n (https://github.com/n8n-io/n8n/security/advisories); advisories published by Flowise (https://github.com/FlowiseAI/Flowise/security/advisories); advisories published by Langflow (https://github.com/langflow-ai/langflow/security/advisories); advisories published by the MCP Python SDK project (https://github.com/modelcontextprotocol/python-sdk/security/advisories); advisories published by MLflow (https://github.com/mlflow/mlflow/security/advisories); CISA Known Exploited Vulnerabilities catalog (CC0). Data as of 2026-11-22. ShieldMap describes how each problem works and how to fix it; it does not publish exploit code. Report a correction to [email protected].