Horizon Alert
Summary of the vulnerability and why it matters
This advisory concerns a vulnerability in Flowise, a tool used for building custom large language model (LLM) flows. The issue could allow an authenticated user to inject and execute malicious JavaScript code on the Flowise server, potentially leading to arbitrary code execution. The main concern is confirming if Flowise is used and if this specific vulnerability is relevant to your environment.
- Code injection in LLM flow building tool.
- Impacts server security if authenticated users exploit it.
- Assess Flowise usage and potential exposure.
Attack Path
How an attacker could exploit the issue
An authenticated user could craft a malicious `baseURL` to inject JavaScript into the Flowise application. This could allow an attacker to escape the sandbox environment and execute arbitrary code on the server.
- Authenticated session required.
- Crafting `baseURL` to inject JavaScript.
- Arbitrary code execution on the server.
Live Threat
Current exploitation, exposure, and threat context
An authenticated user could inject malicious JavaScript into Flowise, allowing them to execute arbitrary code on the server. This could impact system data and potentially sensitive information when the user has an authenticated session.
- Server-side code execution.
- Malicious input bypasses validation.
- Arbitrary code runs on server.
Operational Fix
Recommended remediation, mitigation, and detection steps
This critical vulnerability in Flowise impacts custom-tool components that execute code within the `vm2` sandbox, allowing authenticated users to inject and run arbitrary JavaScript on the Flowise server. Owners of Flowise deployments, likely platform or application teams, must first identify all instances of Flowise, assess their exposure and criticality, and then coordinate remediation.
- Platform/Application teams should own the issue.
- Verify Flowise instances and accessibility first.
- Plan risk-based remediation and vendor coordination.