critical CVSS 9.8 Actively exploited Public exploit available

CVE-2026-39987·Python vulnerability

marimo is a reactive Python notebook. Prior to 0.23.0, Marimo has a Pre-Auth RCE vulnerability. The terminal WebSocket endpoint /terminal/ws lacks authentication validation, allowing an unauthenticated attacker to obtain a full PTY shell and execute arbitrary system commands. Unlike other WebSocket endpoints (e.g., /ws) that correctly call validate_auth() for authentication, the /terminal/ws endpoint only checks the running mode and platform support before accepting connections, completely skipping authentication verification. This vulnerability is fixed in 0.23.0.

Severity
critical
Software
Python
Fixed in
0.23.0
Published
2026-04-09

Affected versions

Until: 0.23.0

Fixed in: 0.23.0

How to fix this CVE

Upgrade Marimo to version 0.23.0 or later. If upgrading is not possible, restrict network access to Marimo instances and ensure they are not exposed to the internet. Block WebSocket connections to /terminal/ws via reverse proxy.

docker pull ghcr.io/marimo-team/marimo:latest
# Verify version >= 0.23.0
# Never expose Marimo directly to the internet without auth proxy

Defensia detects this vulnerability

What an exploitation attempt looks like

Sample log line indicative of exploitation attempts:

GET /terminal/ws HTTP/1.1
Upgrade: websocket
Connection: Upgrade

The attacker connects to the WebSocket endpoint at /terminal/ws without any authentication token or session cookie. Upon successful WebSocket upgrade, they have an interactive terminal with the same privileges as the Marimo server process, enabling arbitrary command execution.

WAF mitigation (if patching is not yet possible)

Add this rule to your WAF to block exploitation attempts while you schedule the patch. 

# Block unauthenticated WebSocket to terminal:
SecRule REQUEST_URI "@streq /terminal/ws" "id:20263998,phase:1,deny,status:403,msg:'CVE-2026-39987 Marimo terminal WebSocket blocked',severity:CRITICAL"

# Nginx:
location /terminal/ws { deny all; return 403; }

How to check if you are affected

  1. Check Marimo version: pip show marimo | grep Version — vulnerable if < 0.23.0
  2. Check for running Marimo instances: ps aux | grep marimo
  3. Check if Marimo is exposed to the network: ss -tlnp | grep marimo
  4. Search access logs for WebSocket upgrade to terminal: grep 'terminal/ws' /var/log/nginx/access.log
  5. Check for outbound connections from Marimo process: lsof -i -P -n | grep marimo

Indicators of compromise

  • WebSocket connections to /terminal/ws from external IPs
  • Unexpected commands executed by the Marimo server process
  • Reverse shell connections originating from Marimo's PID
  • New files or modified system files created by the Marimo user
  • Cryptocurrency miners or malware downloaded via the terminal

FAQ

What is Marimo?

Marimo is a reactive Python notebook that runs as a web application. It is gaining popularity as a modern alternative to Jupyter notebooks, especially for data science and ML workflows. It includes a built-in terminal feature for command execution.

Is this exploitable from the internet?

Yes, if the Marimo instance is exposed to the internet without authentication proxy. The WebSocket endpoint requires no authentication at all. Any attacker who can reach the Marimo port gets a full shell.

What privileges does the attacker get?

The attacker gets a terminal session running as the same user as the Marimo process. If Marimo runs as root (not recommended but sometimes seen), the attacker gets root access. Otherwise, they get the service user's privileges.

Does this affect Jupyter notebooks too?

No. This vulnerability is specific to Marimo's terminal WebSocket implementation. Jupyter has its own authentication mechanisms (tokens) that protect terminal access.

How do I check if I was compromised?

Check process trees for the Marimo process (unexpected child processes), review shell history for the Marimo user, scan for newly created files, and check for unauthorized outbound connections.

Related Python CVEs

CVE-2026-33054CVSS 10Mesop is a Python-based UI framework that allows users to build web applications. Versions 1.2.2 and below contain a Path Traversal vulnerability that allows any user supplying an untrusted state_token through the UI stream payload to arbitrarily target files on the disk under the standard file-based runtime backend. This can result in application denial of service (via crash loops when reading non-msgpack target files as configurations), or arbitrary file manipulation. This vulnerability heavily exposes systems hosted utilizing FileStateSessionBackend. Unauthorized malicious actors could interact with arbitrary payloads overwriting or explicitly removing underlying service resources natively outside the application bounds. This issue has been fixed in version 1.2.3.
CVE-2026-34938CVSS 10PraisonAI is a multi-agent teams system. Prior to version 1.5.90, execute_code() in praisonai-agents runs attacker-controlled Python inside a three-layer sandbox that can be fully bypassed by passing a str subclass with an overridden startswith() method to the _safe_getattr wrapper, achieving arbitrary OS command execution on the host. This issue has been patched in version 1.5.90.
CVE-2026-25632CVSS 10EPyT-Flow is a Python package designed for the easy generation of hydraulic and water quality scenario data of water distribution networks. Prior to 0.16.1, EPyT-Flow’s REST API parses attacker-controlled JSON request bodies using a custom deserializer (my_load_from_json) that supports a type field. When type is present, the deserializer dynamically imports an attacker-specified module/class and instantiates it with attacker-supplied arguments. This allows invoking dangerous classes such as subprocess.Popen, which can lead to OS command execution during JSON parsing. This also affects the loading of JSON files. This vulnerability is fixed in 0.16.1.
CVE-2026-28505CVSS 10Tautulli is a Python based monitoring and tracking tool for Plex Media Server. Prior to version 2.17.0, the str_eval() function in notification_handler.py implements a sandboxed eval() for notification text templates. The sandbox attempts to restrict callable names by inspecting code.co_names of the compiled code object. However, co_names only contains names from the outer code object. When a lambda expression is used, it creates a nested code object whose attribute accesses are stored in code.co_consts, NOT in code.co_names. The sandbox never inspects nested code objects. This issue has been patched in version 2.17.0.
CVE-2026-9135CVSS 9.9IBM Langflow OSS 1.0.0 through 1.10.0 Langflow versions up to 1.9.2 (commit 94981c443d4918517b9e8163d70fc598dc33a32d) contain a code injection vulnerability in the Policies component's ToolGuard integration that bypasses the allow_custom_components=false security control. The vulnerability exists because the validation mechanism only checks the main component source code in node_template["code"]["value"] but fails to validate dynamic CodeInput fields that store generated ToolGuard Python files. Attackers can embed malicious Python code in these unvalidated dynamic fields, which are persisted in Flow.data and later executed server-side when a guarded tool is invoked through the ToolGuard runtime. This allows authenticated users with flow creation privileges to achieve arbitrary Python code execution on the backend despite custom component restrictions. The vulnerability can be escalated through cross-tenant flow manipulation via the agentic MCP update_flow_component_field tool, which accepts attacker-controlled user_id parameters, enabling attackers to inject malicious code into victim users' flows. When combined with publicly accessible flows and specific misconfigurations (AUTO_LOGIN=true, NEW_USER_IS_ACTIVE=true), the attack can be conducted with reduced authentication requirements.

References

Track CVEs across your fleet automatically

Defensia scans your Linux servers and tells you exactly which ones are running vulnerable versions — including CVE-2026-39987. Free for 1 server.

Get started free