PYSEC-2026-4239: Langflow: OS command injection (RCE) via arbitrary command in MCP stdio server configuration
Langflow is a tool for building and deploying AI-powered agents and workflows. Before Langflow 1.10.3, the MCP stdio transport launched whatever command / args a user put in an MCP server configuration, with no allowlist and (before 1.10.3) wrapped in bash -c "exec {command} ...". Any user able to reach the MCP server settings ("Settings → MCP Servers → Add MCP Server", POST/PATCH /api/v2/mcp/servers/{servername}) or to build a flow with the MCP Tools component could add a "server" whose command is an arbitrary OS command (touch, rm -rf, a reverse shell, ...). The command runs on the Langflow host as the Langflow process user as soon as Langflow tries to connect to the server (listing servers, loading tools, running the flow) — even when the UI then reports that the stdio server failed to start. With the default LANGFLOWAUTOLOGIN=true, GET /api/v1/autologin hands out a token without credentials, so on an exposed instance running the default configuration this is reachable without an account. AUTOLOGIN is documented as a development-only setting; with it disabled, any authenticated (non-admin) user can exploit it. This issue is fixed in Langflow 1.10.3, langflow-base 0.10.3, and lfx 1.10.3.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
Langflowto a version that resolves this vulnerability.Fixed in 1.10.3 - Upgrade
Upgrade
langflow-baseto a version that resolves this vulnerability.Fixed in 0.10.3 - Upgrade
Upgrade
lfxto a version that resolves this vulnerability.Fixed in 1.10.3
Event History
Frequently Asked Questions
Who can exploit this issue on a deployed instance?
An unauthenticated attacker can exploit an exposed instance that has the default LANGFLOW_AUTO_LOGIN=true setting, because the auto-login endpoint issues a token without credentials. If auto-login is disabled, any authenticated user, including a non-administrator, can exploit it.
What access or action is required to trigger command execution?
The attacker must be able to create or modify an MCP server configuration through the MCP server settings or the POST/PATCH MCP server API, or build a flow using the MCP Tools component. The configured command executes when Langflow attempts to connect to that server, such as while listing servers, loading tools, or running the flow.
Does a failed MCP stdio server startup prevent exploitation?
No. The command can run on the Langflow host as the Langflow process user even if the Langflow UI subsequently reports that the stdio server failed to start.
What versions contain the fix?
The issue is fixed in Langflow 1.10.3 and langflow-base 0.10.3. The provided data does not state a fixed lfx version.