Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessAuthenticate to n8n with low-privileged account
-
DeliveryIdentify AI agent workflow using computer-use package
-
ExploitCraft traversal pattern to bypass base-directory check
-
ExecutionSubmit pattern to search_files tool
-
PersistTool resolves paths outside configured sandbox
-
ImpactRetrieve arbitrary file contents readable by daemon user
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A01:2025 - Broken Access Control
ASVS 5.0
V5 - File Handling Path canonicalization, traversal prevention
MITRE ATT&CK
2 techniques across 2 tactics: Discovery, Collection The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up. |
| Exploitation Conditions | Exploitation requires three conditions to be simultaneously true. First, the actor must hold a low-privileged authenticated account on the n8n instance, as confirmed by PR:L in the CVSS 4.0 vector - unauthenticated external attackers cannot exploit this directly. Second, the targeted n8n deployment must have AI agent workflows configured that use the @n8n/computer-use package with the search_files tool active; this is a non-default component not present in standard n8n installations, and the AT:P in the CVSS 4.0 vector explicitly captures this prerequisite. Third, the actor must be able to control or influence the search pattern input passed to the search_files tool within such a workflow. Exploitation impact is further bounded by the OS-level read permissions of the n8n daemon's process account; running n8n under a dedicated low-privilege user with restricted home directory access materially limits accessible file paths even if the confinement bypass succeeds. |
| Impact | The CVSS 4.0 vector (AV:N/AC:L/AT:P/PR:L/UI:N/VC:N/VI:N/VA:N/SC:H/SI:N/SA:N) produces a score of 4.9 (Medium), with the high subsequent-system confidentiality impact (SC:H) as the sole driver - reflecting the cross-boundary arbitrary file read - offset by the AT:P (Attack Requirements: Present) constraint. That AT:P metric is the critical limiting factor: the @n8n/computer-use package and an AI agent workflow using search_files must both be deployed and accessible, which is not a default configuration. PR:L confirms that some authenticated access to the n8n instance is required, ruling out unauthenticated external exploitation. No CISA KEV listing exists and no public proof-of-concept has been identified, further reducing operational urgency relative to actively-exploited path traversal flaws. However, the gap between the modest numeric score and the real-world impact (SC:H equates to reading arbitrary host files) is meaningful in multi-tenant or SaaS-style n8n instances: a low-privileged user could pivot to read database credentials, API keys, or other secrets stored on the host, making this a credible lateral-movement vector. Prioritization should be driven by whether computer-use AI workflows are active in a given deployment. |
| Exploit ScenarioDerived | An authenticated n8n user with low privileges locates an AI agent workflow in the instance that exposes the computer-use search_files tool. The attacker submits a crafted search pattern - for example a glob expression containing relative traversal components - that circumvents the base-directory confinement check and resolves to paths on the host filesystem. The tool returns file names and contents from outside the sandbox, such as n8n's own .env file containing database connection strings or third-party API keys, which the attacker harvests for further account compromise or lateral movement. No public proof-of-concept code has been identified, but path traversal via crafted glob patterns is a well-understood and straightforward attack primitive. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Aug 11, 2026
Path traversal in n8n's @n8n/computer-use search_files tool exposes arbitrary local files to any actor who can control the tool's search input. Affected versions span all n8n releases before 2.31.5 and the entire 2.32.x branch before 2.32.1, limited to deployments that run AI agent workflows leveraging the computer-use package. An authenticated low-privileged user can bypass the configured base-directory sandbox and retrieve the names and contents of any file readable by the n8n daemon's OS user account. No public exploit code or CISA KEV listing has been identified at the time of analysis.
RemediationAI
Vendor-released patches are confirmed by the GitHub GHSA advisory (GHSA-pf2q-pxhf-hgmw): upgrade n8n to version 2.31.5 on the stable branch or 2.32.1 on the 2.32.x branch. Both fix versions are independently confirmed via GHSA npm package vulnerability data and represent the primary remediation. For deployments where immediate upgrade is not feasible, the vendor recommends three temporary mitigations. First, restrict n8n instance access to fully trusted users only - this reduces but does not eliminate risk in multi-user environments and has no technical side effects. Second, disable or remove all AI agent workflows that invoke the @n8n/computer-use package; this directly eliminates the vulnerable attack surface and is the most effective short-term control, with the trade-off of losing AI computer-use functionality. Third, ensure the n8n process runs under a dedicated low-privilege OS user account with minimal read access - avoiding access to paths such as /etc, /root, or application credential directories - which narrows the blast radius of exploitation without removing functionality. None of these workarounds fully remediate the underlying path traversal flaw and the vendor explicitly notes they are short-term measures only.
Affected VersionsENISA
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
n8n is a Node.js-based workflow automation platform distributed via npm (CPE: cpe:2.3:a:n8n-io:n8n). The vulnerable component is the @n8n/computer-use package's file-search (search_files) tool, which is invoked from within AI agent workflows. The tool is designed to confine file-system queries to an operator-configured base directory acting as a security sandbox. The root cause is CWE-22 (Path Traversal): the confinement check fails to canonicalize or validate search patterns before applying access-control decisions, allowing crafted inputs - such as glob expressions with relative traversal components - to expand to paths outside the permitted tree. Because the sandbox escape is at the OS level, the breadth of disclosure is bounded only by the read permissions of the n8n daemon's process account, potentially exposing environment files, application secrets, host credentials, or other sensitive configuration on the underlying host.
Know when N8n 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:P/PR:L/UI:N/VC:N/VI:N/VA:N/SC:H/SI:N/SA:N
PR:L for required login; AC:H maps the AT:P condition that computer-use workflows must be deployed and actor controls search input; S:C for sandbox boundary breach; C:H for arbitrary host file read; no integrity or availability impact.
Primary rating from Vendor (VulnCheck).
CVSS VectorVendor: VulnCheck
Lifecycle Timeline
1DescriptionCVE.org
n8n before 2.31.5 and 2.32.x before 2.32.1 contain a path-confinement bypass in the @n8n/computer-use file-search (search_files) tool. A crafted search pattern can bypass the base-directory confinement check and expand to locations outside the configured directory, causing the tool to return the names and contents of arbitrary local files readable by the daemon's OS user. Any deployment where an actor can influence the tool's search input is affected.
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
Path traversal exploits occur when applications use user-controlled input to construct file system paths without proper validation.
Remote code execution in n8n's workflow expression evaluation system allows an authenticated user with workflow creation
n8n workflow automation (through 1.121.2) allows authenticated users to execute arbitrary code via the n8n service, with
n8n workflow automation (1.65.0 to 1.121.0) allows unauthenticated file access through form-based workflows. A critical
n8n has a fifth critical RCE vulnerability (CVSS 9.9) in the Expression evaluator, enabling code execution through craft
Authenticated users who can create or modify an n8n workflow can achieve remote code execution on the n8n host by abusin
A prototype pollution vulnerability in the XML and GSuiteAdmin nodes of n8n workflow automation platform allows authenti
An authenticated user with workflow creation or modification privileges in n8n workflow automation platform can exploit
n8n-workflows Main Commit ee25413 allows attackers to execute a directory traversal via the download_workflow function w
Authenticated n8n users can hijack administrator accounts when LDAP authentication is enabled by manipulating their LDAP
The n8n package 0.218.0 for Node.js allows Escalation of Privileges. Rated high severity (CVSS 8.8), this vulnerability
SQL injection in n8n's Data Table Get node allows authenticated users with workflow modification permissions to execute
Authenticated users can exploit string formatting and exception handling in n8n's Python task executor to escape sandbox
EUVD-2026-56099
GHSA-cgh8-xfc3-8rm5
NVD
CVE.org
EPSS
Vulnrichment
VulDB
Exploit-DB