Flowise
CVE-2025-34267
HIGH
Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessAuthenticate to Flowise with tool-authoring rights
-
DeliveryCreate Custom Tool using Puppeteer/Playwright
-
ExploitSet attacker-controlled browser binary path and args
-
ExecutionRun tool to spawn executable on host
-
PersistEscape nodevm sandbox
-
ImpactExecute arbitrary code in host context
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A05:2025 - Injection
ASVS 5.0
V1 - Encoding and Sanitization Command injection prevention
MITRE ATT&CK
1 technique across 1 tactic: Execution The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up. |
| Exploitation Conditions | Requires an authenticated Flowise account with privileges to create or run a Custom Tool/Function (CVSS PR:H), and the account must be able to invoke the integrated Puppeteer or Playwright modules from within the nodevm execution environment. On versions 3.0.1 through pre-3.0.8 this path is reachable by default; on 3.0.8 and later it is only exploitable when the operator has explicitly set ALLOW_BUILTIN_DEP=true, since the fix gates built-in/project dependency exposure behind that flag (default false). Exploitation is limited by that required privilege level and, on patched builds, by the non-default configuration toggle; no end-user interaction beyond the attacker's own tool execution is needed (UI:N). |
| Impact | The CVSS 4.0 base score of 8.4 (High) with vector AV:N/AC:L/AT:N/PR:H/UI:N and VC:H plus subsequent-system impact (SC:H/SI:H) reflects network reachability, low complexity, no user interaction, but a requirement for high privileges (PR:H) - i.e. an authenticated account with rights to create or run tools. That authentication requirement, combined with the non-default ALLOW_BUILTIN_DEP toggle needed for versions after the fix, materially limits mass exploitation despite the network attack vector. Publicly available exploit code exists, which raises practical risk for exposed or multi-tenant Flowise instances, but there is no CISA KEV listing and no EPSS score was provided in the input, so widespread automated exploitation is not evidenced - this reads as a targeted, post-authentication escalation-to-host risk rather than an internet-wide worming threat. Net priority is High for any organization running affected Flowise versions with untrusted or low-trust users who can author tools, and elevated where the instance is internet-exposed; lower where Flowise is single-tenant and access is tightly controlled. |
| Exploit ScenarioDerived | An authenticated Flowise user with tool-creation rights builds a Custom Tool that invokes the bundled Puppeteer or Playwright API, pointing the browser executablePath at an attacker-chosen binary (or supplying malicious launch arguments). When the tool runs, Flowise spawns that executable on the host outside the nodevm sandbox, giving the attacker arbitrary command execution in the host context. Public exploit code exists, and given AC:L/UI:N the tool executes reliably once created by a privileged account. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Jul 14, 2026
Authenticated remote code execution in FlowiseAI Flowise (v3.0.1 up to but not including 3.0.8, and later versions when 'ALLOW_BUILTIN_DEP' is enabled) lets a logged-in user break out of the nodevm sandbox by abusing the bundled Puppeteer and Playwright modules. Because a custom tool can specify an attacker-controlled browser binary path and launch parameters, executing that tool runs an arbitrary executable on the host outside the intended sandbox, yielding code execution in the host context. Publicly available exploit code exists; the flaw is not listed in CISA KEV, and it was mis-filed by the vendor as a duplicate of CVE-2025-26319 but is distinct.
RemediationAI
Vendor-released patch: upgrade to Flowise 3.0.8 or later, which changes packages/components/src/utils.ts so that built-in/project dependencies are only exposed to the sandbox when ALLOW_BUILTIN_DEP is explicitly 'true' (fix in https://github.com/FlowiseAI/Flowise/pull/5231, advisory https://github.com/FlowiseAI/Flowise/security/advisories/GHSA-5w3r-f6gm-c25w). Because the setting still enables the dangerous path on patched versions, keep ALLOW_BUILTIN_DEP unset or false (the new default) unless you fully trust every user who can author tools; setting it to true re-introduces host code execution and should be avoided in shared or exposed deployments. If you cannot upgrade immediately, restrict who holds tool-creation/execution privileges, avoid exposing the Flowise UI/API to untrusted networks, and run Flowise as an unprivileged, containerized/sandboxed OS user so a host escape is contained - noting these controls reduce blast radius but do not remove the vulnerability, which the version upgrade does.
Affected Packages
Source: OSV (Open Source Vulnerabilities)
Affects your stack
Affects 1 package in npm
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
Flowise is an open-source low-code/no-code visual builder for LLM orchestration and AI agent workflows. Its Custom Tool / Function feature runs user-supplied JavaScript inside a Node 'vm' (nodevm) sandbox that is meant to restrict access to host resources and dependencies. The bundled browser-automation libraries Puppeteer and Playwright expose an interface to launch an external browser process by explicit executable path and arguments; when the integrated modules are made available to sandboxed code (gated by ALLOW_BUILTIN_DEP), that launch capability becomes a primitive to spawn an arbitrary host process, defeating the sandbox. This maps to CWE-77 (command injection / improper neutralization of special elements in a command), the root cause being that attacker-influenced executable paths and parameters are passed to a process-spawning API without restriction. The PR #5231 fix confirms the mechanism: in packages/components/src/utils.ts the available built-in/project dependencies are now only concatenated into the sandbox dependency set when process.env.ALLOW_BUILTIN_DEP === 'true', otherwise only explicitly declared external deps are exposed, and ALLOW_BUILTIN_DEP is introduced across env examples and docker-compose files defaulting to false.
Know when Flowise gets its next vulnerability
- Watch products and get an email when something new is published
- Check this CVE against your own stack and servers
- Exploit and PoC references, threat intelligence from 22 sources
No credit card · 7-day full trial
Severity by source
CVSS:4.0/AV:N/AC:L/AT:N/PR:H/UI:N/VC:H/VI:L/VA:L/SC:H/SI:H/SA:L
Network-reachable tool execution needs an authenticated privileged account (PR:H) but no interaction; sandbox escape to host is a scope change (S:C) with full host code execution (C/I/A:H).
Primary rating from Vendor (vulncheck).
CVSS VectorVendor: vulncheck
DescriptionCVE.org
Flowise v3.0.1 < 3.0.8 and all versions after with 'ALLOW_BUILTIN_DEP' enabled contain an authenticated remote code execution vulnerability and node VM sandbox escape due to insecure use of integrated modules (Puppeteer and Playwright) within the nodevm execution environment. An authenticated attacker able to create or run a tool that leverages Puppeteer/Playwright can specify attacker-controlled browser binary paths and parameters. When the tool executes, the attacker-controlled executable/parameters are run on the host and circumvent the intended nodevm sandbox restrictions, resulting in execution of arbitrary code in the context of the host. This vulnerability was incorrectly assigned as a duplicate CVE-2025-26319 by the developers and should be considered distinct from that identifier.
Executive ViewAI
FlowiseAI Flowise versions 3.0.1-3.0.7 (and later versions with ALLOW_BUILTIN_DEP enabled) contain an authenticated remote code execution vulnerability with publicly available exploit code. Logged-in users can execute arbitrary system commands outside the application sandbox, with code executing in the host system context where Flowise runs-creating a direct path to full system compromise for any organization running affected versions.
No vendor-released patch identified at time of analysis. Interim mitigations include: disable ALLOW_BUILTIN_DEP environment variable; restrict Flowise UI access to VPN and internal networks only, limiting user accounts to essential administrative staff; disable or prevent users from creating custom tools leveraging Puppeteer or Playwright; run Flowise in a restrictive container with minimal host filesystem and process permissions; implement detailed logging and alerting on all tool executions, particularly browser automation invocations; evaluate downgrade to pre-3.0.1 versions as interim measure; and maintain vendor monitoring for patch availability.
Threat Intelligence Overview
How this decision was reached
The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.
What every action level means
SSVC AssessmentENISA
Priority Score
Attack Technique
OS command injection occurs when an application passes unsanitized user input directly into system shell commands.
Remote Code Execution represents the critical moment when an attacker successfully runs arbitrary code on a target system without physical access.
Affected Products (CPE)CVE.org
Flowise version 3.0.5 contains a remote code execution vulnerability in the CustomMCP node. The mcpServerConfig paramete
Flowise versions before 3.0.1 allow unauthenticated access to the Custom MCPs feature, which is designed to execute OS c
FlowiseAI Flowise version 2.2.6 contains an arbitrary file upload vulnerability in the /api/v1/attachments endpoint. Una
Arbitrary file read and write in Flowise versions prior to 3.0.8 allows an authenticated attacker with access to the flo
Remote code execution in Flowise before 3.1.2 allows any authenticated user (or API caller with chatflow view/update per
Flowise is a drag & drop user interface to build a customized large language model flow. Rated critical severity (CVSS 9
Authenticated users can upload arbitrary files to Flowise 3.0.7 due to missing validation, enabling persistent storage o
Unrestricted file upload in Flowise LLM workflow builder before 3.0.13 via /api/v1/attachments endpoint allows unauthent
Missing authentication on NVD data endpoint in Flowise before 3.0.13 allows unauthenticated access to internal vulnerabi
Remote code execution in Flowise before 3.1.3 arises because its Pyodide Python-code validator (validatePythonCodeForDat
Remote code execution in FlowiseAI Flowise before 3.1.3 lets an attacker execute arbitrary OS commands as the Flowise pr
Remote command execution in Flowise <= 3.1.2 lets an authenticated user break out of the vm2/@flowiseai/nodevm JavaScrip
NVD
CVE.org
EPSS
GitHub
VulDB
Exploit-DB