Skip to main content

browser-use web-ui CVE-2026-82637

| EUVDEUVD-2026-68213 MEDIUM
External Control of File Name or Path (CWE-73)
2026-08-30 VulnCheck GHSA-hhqw-gcq9-6cmv
6.9
CVSS 4.0 · Vendor: VulnCheck
View

Attack ChainAIDerived

Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain

  1. Access
    Access unauthenticated Gradio interface
  2. Delivery
    Submit crafted absolute path in save_*_path parameter
  3. Exploit
    run_agent_task passes path unvalidated to makedirs
  4. Execution
    Container creates arbitrary directory as root
  5. Impact
    Leverage sensitive directory path for persistence or privilege escalation

Vulnerability Assessment

Weakness
CWE
OWASP Top 10:2025
A06:2025 - Insecure Design
ASVS 5.0
V5 - File Handling File name/path control
Exploitation Conditions

Exploitation requires network access to the Gradio web interface of browser-use web-ui, typically served on port 7860/tcp. No authentication is required - the Gradio interface is unauthenticated by default, confirmed by PR:N in the CVSS 4.0 vector. The container must be running as root (the documented default deployment posture) for the broadest impact; non-root container deployments limit exploitable write targets but do not eliminate the path injection primitive entirely. No prior session state, browser interaction, or special protocol knowledge is required - a single crafted HTTP POST to the agent task endpoint is sufficient. Exploitation impact is bounded to directory creation (not arbitrary file write or read), so chained impact depends on whether the attacker can leverage created directory paths in privileged contexts (e.g., cron drop-ins, sudoers fragments, or dynamic-library search paths). Deployments exposing the Gradio interface to untrusted or internet-facing networks are the highest-risk population.

Impact

The CVSS 4.0 score of 6.9 with vector AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:L/VA:N reflects low-complexity, unauthenticated, network-reachable exploitation with a low integrity-only impact - a reasonable characterization of the isolated primitive (directory creation, not arbitrary file write). EPSS data is absent from the input, preventing exploitation-probability quantification. The vulnerability is not in CISA KEV and no public exploit has been identified at time of analysis. Despite the moderate score, real-world risk is amplified by two compounding factors: the Gradio interface has no authentication by default (confirmed PR:N), and the container's root execution context means an attacker can create directories in sensitive locations - such as /etc/cron.d, /etc/sudoers.d, or /usr/local/bin - that could serve as stepping stones to persistence or privilege escalation beyond what VI:L alone implies. This is not a high-severity exploit primitive in isolation, but the zero-auth network exposure combined with root-context execution warrants prompt patching rather than reliance on compensating controls alone.

Exploit ScenarioDerived

An attacker with network access to the exposed Gradio interface sends a crafted request to the run_agent_task endpoint, supplying an absolute path such as /etc/cron.d/implant as the value of save_recording_path. The server creates the directory without path validation because it executes as root within the container. In environments where the container filesystem is bind-mounted to the host, this primitive could be chained to plant directories in host-sensitive locations, enabling persistence or privilege escalation.

This Month Plan within 30 days. Why: MEDIUM severity (CVSS 4.0-6.9).
No patch listed yet 0.2% exploitation probability (EPSS)

Get an e-mail when this CVE is exploited, patched or re-scored. Free account.

Alert me

AnalysisAI

Arbitrary directory creation in browser-use web-ui 2.0.0-3.0.0 allows remote unauthenticated attackers to create directories at any path writable by the root-running container. The run_agent_task function passes user-supplied save_recording_path, save_trace_path, save_agent_history_path, and save_download_path values directly to filesystem operations without sanitizing absolute path inputs. Because the Gradio interface requires no authentication and containers commonly run as root, the effective write surface spans the entire container filesystem. No public exploit has been identified at time of analysis.

RemediationAI

Upgrade browser-use web-ui to a version beyond 3.0.0 once a patched release is available; no specific patched version number is confirmed in the available data - track the GitHub repository at https://github.com/browser-use/web-ui and the issue at https://github.com/browser-use/web-ui/issues/737 for patch status. The upstream fix should validate that all four path parameters (save_recording_path, save_trace_path, save_agent_history_path, save_download_path) are relative paths or resolve strictly within an allow-listed base directory before any directory-creation call. As an immediate compensating control, restrict network access to the Gradio interface (typically port 7860/tcp) to trusted IP ranges only via firewall or reverse-proxy allowlist - this removes the unauthenticated remote attack vector at the cost of blocking legitimate remote access. A second compensating control is running the container as a non-root user, which limits which host or container paths an attacker can target even if exploited, though it does not eliminate the path injection primitive itself.

Affected VersionsENISA

web-ui 2.0.0 ≤3.0.0

Affected Packages

euvd web_ui no fix

Source: OSV (Open Source Vulnerabilities)

Affects your stack

Affects 1 package in euvd

Checking this CVE against your own stacks and servers is part of the Basic plan and up.

Technical ContextAI · show details

browser-use web-ui is a Python-based Gradio frontend (cpe:2.3:a:browser-use:web-ui:*:*:*:*:*:*:*:*) for the browser-use AI browser automation library, covering versions 2.0.0 through 3.0.0. The vulnerability resides in src/webui/components/browser_use_agent_tab.py within the run_agent_task function, where four path parameters accepted from the Gradio UI are passed to directory-creation calls without stripping absolute prefixes or confining paths to an intended base directory. CWE-73 (External Control of File Name or Path) describes this root cause precisely: user-controlled data flows into a filesystem path constructor without sanitization, enabling path injection. The Gradio interface is served unauthenticated by default, and the container is documented to run as root, eliminating typical filesystem permission barriers that would otherwise constrain the blast radius of low-integrity write primitives.

Know when Web Ui 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

Vendor (VulnCheck) PRIMARY 6.9 MEDIUM
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N
vuln.today AI 5.3 MEDIUM

Unauthenticated network access via default Gradio interface (AV:N, PR:N); impact limited to low-integrity directory creation with no confidentiality or availability effect.

3.1 AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N
4.0 AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N

Primary rating from Vendor (VulnCheck).

CVSS VectorVendor: VulnCheck

Attack Vector
Network
Attack Complexity
Low
Attack Requirements
None
Privileges Required
None
User Interaction
None
Vulnerable system: Confidentiality
None
Vulnerable system: Integrity
Low
Vulnerable system: Availability
None
Subsequent system: Confidentiality
None
Subsequent system: Integrity
None
Subsequent system: Availability
None

DescriptionCVE.org

browser-use web-ui versions 2.0.0 through 3.0.0 fail to validate browser settings paths in run_agent_task, allowing attackers to create directories at arbitrary locations by supplying absolute paths to save_recording_path, save_trace_path, save_agent_history_path, or save_download_path parameters. Attackers can exploit this via the unauthenticated Gradio interface to create directories anywhere the root-running container has write access.

Threat Intelligence Overview

EPSS Score
0.2%
Top 85% exploitable
Prevalence
Low
niche AI browser-automation developer tool, limited enterprise production footprint
Data Quality
5 sources
Confidence: Medium

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.

SignalValue
Severity MEDIUM
CISA KEV No
Public POC No
EPSS 0.2%
Patch available No
Active threat No
Deciding rule
MEDIUM severity (CVSS 4.0-6.9)
What every action level means
Emergency Immediate action required. Criteria: CRITICAL + in CISA KEV (actively exploited) without patch, OR CRITICAL + EPSS > 30% without patch, OR CRITICAL + active threat signals without patch.
Act Now Prioritize within 24-48h. Criteria: in KEV with patch available, OR CRITICAL + public POC, OR HIGH + EPSS > 10%, OR CRITICAL + active threat, OR any CRITICAL severity.
This Week Schedule remediation this week. Criteria: HIGH severity (CVSS 7.0-8.9). Patch within normal change window.
This Month Plan within 30 days. Criteria: MEDIUM severity (CVSS 4.0-6.9), lower exploitation probability. Include in regular patch cycle.
Monitor Track and reassess. Criteria: LOW severity (CVSS 0.1-3.9), minimal exploitation risk. No immediate action needed.

SSVC AssessmentENISA

SSVC Decision· derived
Scheduled
Exploitation
Poc
System Exposure· derived
Open
Automatable
Yes
Human Impact· derived
Medium
Technical Impact
Partial

Priority Score

34
Low Medium High Critical
KEV: 0
EPSS: 0
CVSS: +34
PoC: 0
Weaponized: 0
Threat intel: 0
Gov advisory: 0

Attack Technique

Information disclosure occurs when an application unintentionally exposes sensitive data that aids attackers in reconnaissance or directly compromises security.

This site uses cookies essential for authentication and security. No tracking or analytics cookies are used. Privacy Policy