External risk intelligence

Flowise CSVAgent Python Injection Command Execution.

CVE advisorySeverity: CRITICAL (CVSS 9.2)

CVE-2026-69255

Flowise is a web-based application designed for building LLM flows and interfaces. These platforms are commonly deployed as internet-facing web services or APIs to allow users or integrated systems to interact with the LLM workflows, making the application's interface and underlying processing components frequently accessible via the network.

Code Injection

Halo Surface Signal: 4 out of 5 — likely to be public-facing.

External exposure likelihood

Horizon Alert

Summary of the vulnerability and why it matters

A vulnerability in Flowise, a tool for building large language model flows, could allow an attacker to execute arbitrary operating system commands as root within the application's container. This occurs because the system improperly handles CSV data, allowing for code injection before security checks are applied. The main concern is confirming relevance and exposure due to the potential for elevated system access.

  • Unvalidated CSV data allows command execution.
  • Root-level access creates a significant security risk.
  • Confirm if this system is in use and needs attention.

Attack Path

How an attacker could exploit the issue

An attacker could exploit this vulnerability by sending specially crafted CSV data to a vulnerable Flowise instance. The system then processes this data in a way that allows the attacker to insert and execute arbitrary Python code. This code can leverage a bridge to Node.js, enabling the execution of operating system commands with root privileges within the Flowise container.

  • No authentication required.
  • Inject malicious code via CSV data.
  • Execute arbitrary commands as root.

Live Threat

Current exploitation, exposure, and threat context

Flowise, a tool for building large language model flows, has a vulnerability where an authenticated attacker could inject and execute arbitrary Python code. This could lead to the execution of operating system commands with root privileges within the Flowise container when supported by the advisory's conditions.

  • Arbitrary code execution within the container.
  • Injection through specially crafted CSV data.
  • Compromise of the containerized environment.

Operational Fix

Recommended remediation, mitigation, and detection steps

System owners and application teams are likely responsible for addressing this vulnerability in Flowise, as it affects a web-based application used for building LLM flows. The first practical step is to identify all instances of Flowise, determine their reachability and criticality, assign an owner, and then plan remediation based on the assessed risk.

  • Application owners should investigate Flowise instances.
  • Verify network exposure and business criticality.
  • Plan remediation, considering vendor coordination.

Supplementary metadata

Validate whether this threat affects your internet-facing exposure.

Halo Threat Intelligence helps prioritize remediation with Halo Surface Signal and H/A/L/O context. Start exposure validation with a free external attack surface trial.

Frequently asked questions

What is Flowise?

Flowise is a drag-and-drop platform used to design and build custom workflows for large language models. It provides a visual interface that allows developers to create complex AI integrations and automated LLM-based processes without needing to write extensive low-level orchestration code.

How does CVE-2026-69255 cause command execution?

This vulnerability is an instance of Improper Control of Generation of Code, or CWE-94. The CSVAgent component processes CSV data and embeds it directly into a Python script without sufficient sanitization. By injecting specific syntax, an attacker can break out of the intended data structure, execute arbitrary Python code, and bridge into the underlying system to run OS commands.

Do I need to worry if I am not using the CSVAgent?

If your Flowise deployment does not utilize the CSVAgent component, you may not be subject to this specific injection path. The vulnerability specifically requires the processing of attacker-provided CSV data through this agent to reach the unsafe code interpolation. If this node is inactive or unused in your LLM flows, the identified trigger path is not met.

Is my Flowise instance at risk from the internet?

According to Halo Surface Signal, Flowise is commonly deployed as an internet-facing web service to allow external systems or users to interact with LLM workflows. Because of this, instances exposed to the network are more easily reachable by attackers looking to send crafted CSV data. Internal-only deployments have a smaller attack surface but still carry risk if accessible to untrusted internal actors.

What is the first step to remediate this vulnerability?

The immediate priority is to upgrade your Flowise installation to version 3.1.3 or later, which contains the security fix. Before applying updates, identify all active Flowise instances in your environment, assess their current network exposure, and verify if they are configured to use the affected CSVAgent node.

References