CVE-2026-103245: n8n: missing Webflow HMAC check allows forged triggers

MEDIUM
Published October 1, 2026
CISO Take

n8n's Webflow Trigger node never verifies the x-webflow-signature HMAC, so anyone who can reach the webhook URL can send forged requests that start workflows with attacker-controlled payloads. This is a medium-severity flaw (CVSS 5.3, integrity impact only). It is not in CISA KEV, no public exploit or Nuclei template exists, and CISA SSVC rates it TRACK. The raw EPSS score is low (0.00157), although the feed places it in a high percentile. The real risk is the blast radius: n8n workflows often feed AI agents, CRMs, databases and third-party APIs, so a forged event can drive record creation or API calls with the workflow's own credentials. Upgrade to n8n 1.123.80, 2.39.6 or 2.40.1 or later, depending on your release line. Until then, restrict network access to webhook endpoints and review the Webflow-triggered workflows for downstream actions with side effects.

Sources:NVDEPSSGitHub AdvisoryVulnCheckATLAS

What is the risk?

Overall risk is MEDIUM. Exploitation needs no authentication, privileges or user interaction (AV:N/AC:L/PR:N/UI:N), but the attacker must know or discover the webhook URL, which contains an unguessable path segment in default n8n deployments. Confidentiality and availability are not directly affected. Integrity impact is limited to what the triggered workflow does, and that can be significant when the workflow has write access to CRMs, databases, payment tools or AI agents with tool permissions. There is no known exploitation in the wild (not in KEV, no public PoC), so urgency is routine-patch unless the affected workflows have high-impact downstream actions. Confidence: HIGH on the facts from the vendor advisory, MEDIUM on real-world exposure, since Unknown downstream dependents and no scorecard data were available.

How does the attack unfold?

Discovery
Attacker obtains or enumerates the n8n Webflow webhook URL from leaked logs, configs or a publicly exposed instance.
AML.T0006
Initial Access
Attacker sends a forged Webflow-style POST without a valid x-webflow-signature, and the handler accepts it.
AML.T0049
Workflow Execution
The forged payload starts the workflow, which runs downstream nodes and AI agent tools with its stored credentials.
AML.T0053
Impact
Attacker-controlled data creates or alters records, triggers unauthorized API calls, or injects instructions into AI agent steps.
AML.T0048

What systems are affected?

PackageEcosystemVulnerable RangePatched
n8nnpm—No patch
210.5KOpenSSF 7.116 dependentsPushed 3d ago42% patched~19d to patchFull package profile →

Do you use n8n? You're affected.

How severe is it?

CVSS 3.1
5.3 / 10
EPSS
0.2%
chance of exploitation in 30 days
Higher than 5% of all CVEs
Exploitation Status
No known exploitation
Sophistication
Trivial

What is the attack surface?

AVACPRUISCIA
AVNetwork
ACLow
PRNone
UINone
SUnchanged
CNone
ILow
ANone

What should I do?

7 steps
  1. Patch to n8n 1.123.80, 2.39.6 or 2.40.1 or later, according to your release line.

  2. If you cannot patch right away, disable or deactivate workflows that use the Webflow Trigger node, or put the webhook path behind a reverse proxy or WAF that allows only Webflow's published source IPs and validates the signature at the edge.

  3. Rotate the webhook URLs/paths of existing Webflow triggers and re-register them in Webflow.

  4. Add validation inside workflows (for example, check payload fields and site IDs against expected values) before any write or API-call step.

  5. Review n8n execution logs for Webflow-triggered runs from unexpected source IPs or with malformed or unusual payloads since the exposure began.

  6. Apply least privilege to the credentials used by these workflows and require human approval for destructive or high-value actions.

  7. Track GHSA-hwv9-jhc7-f7c4 and the VulnCheck advisory for updates.

What does CISA's SSVC say?

DecisionTrack
Exploitationnone
AutomatableYes
Technical Impactpartial

Source: CISA Vulnrichment (SSVC v2.0). Decision based on the CISA Coordinator decision tree.

How is it classified?

Which compliance frameworks are affected?

This CVE is relevant to:

EU AI Act
Article 15-Accuracy, robustness and cybersecurity
ISO 42001
A.6.2.6-AI system operation and monitoring
NIST AI RMF
MEASURE 2.7-AI system security and resilience are evaluated and documented
OWASP LLM Top 10
LLM06-Excessive Agency

Frequently Asked Questions

What is CVE-2026-103245?

n8n's Webflow Trigger node never verifies the x-webflow-signature HMAC, so anyone who can reach the webhook URL can send forged requests that start workflows with attacker-controlled payloads. This is a medium-severity flaw (CVSS 5.3, integrity impact only). It is not in CISA KEV, no public exploit or Nuclei template exists, and CISA SSVC rates it TRACK. The raw EPSS score is low (0.00157), although the feed places it in a high percentile. The real risk is the blast radius: n8n workflows often feed AI agents, CRMs, databases and third-party APIs, so a forged event can drive record creation or API calls with the workflow's own credentials. Upgrade to n8n 1.123.80, 2.39.6 or 2.40.1 or later, depending on your release line. Until then, restrict network access to webhook endpoints and review the Webflow-triggered workflows for downstream actions with side effects.

Is CVE-2026-103245 actively exploited?

No confirmed active exploitation of CVE-2026-103245 has been reported, but organizations should still patch proactively.

How to fix CVE-2026-103245?

1. Patch to n8n 1.123.80, 2.39.6 or 2.40.1 or later, according to your release line. 2. If you cannot patch right away, disable or deactivate workflows that use the Webflow Trigger node, or put the webhook path behind a reverse proxy or WAF that allows only Webflow's published source IPs and validates the signature at the edge. 3. Rotate the webhook URLs/paths of existing Webflow triggers and re-register them in Webflow. 4. Add validation inside workflows (for example, check payload fields and site IDs against expected values) before any write or API-call step. 5. Review n8n execution logs for Webflow-triggered runs from unexpected source IPs or with malformed or unusual payloads since the exposure began. 6. Apply least privilege to the credentials used by these workflows and require human approval for destructive or high-value actions. 7. Track GHSA-hwv9-jhc7-f7c4 and the VulnCheck advisory for updates.

What systems are affected by CVE-2026-103245?

This vulnerability affects the following AI/ML architecture patterns: Agent frameworks, Workflow automation / orchestration, Webhook-triggered AI pipelines, LLM agent tool chains.

What is the CVSS score for CVE-2026-103245?

CVE-2026-103245 has a CVSS v3.1 base score of 5.3 (MEDIUM). The EPSS exploitation probability is 0.16%.

What is the AI security impact?

Affected AI Architectures

Agent frameworksWorkflow automation / orchestrationWebhook-triggered AI pipelinesLLM agent tool chains

MITRE ATLAS Techniques

AML.T0049Exploit Public-Facing Application
AML.T0053AI Agent Tool Invocation
AML.T0093Prompt Infiltration via Public-Facing Application

Compliance Controls Affected

EU AI Act: Article 15
ISO 42001: A.6.2.6
NIST AI RMF: MEASURE 2.7
OWASP LLM Top 10: LLM06

What are the technical details?

Original Advisory

n8n versions before 1.123.80, from 2.0.0 before 2.39.6, and from 2.40.0 before 2.40.1 fail to verify the x-webflow-signature HMAC in the Webflow Trigger node webhook handler. Unauthenticated attackers can send forged webhook requests with attacker-controlled payloads to trigger workflows and manipulate downstream actions like record creation or API calls.

Exploitation Scenario

An attacker finds a company's n8n Webflow webhook URL through leaked logs, a shared screenshot, a misconfigured proxy, or by enumerating a public instance. They send an HTTP POST that mimics a Webflow form-submission event, with no valid x-webflow-signature, and n8n accepts it. The workflow processes the payload as a trusted lead: it creates a CRM record, calls billing or messaging APIs, and passes the free-text fields to an LLM agent node. The attacker embeds instructions in those fields to steer the agent's tool calls, or simply floods the system with fake records to corrupt data and trigger costly API calls.

Weaknesses (CWE)

CWE-347 — Improper Verification of Cryptographic Signature: The product does not verify, or incorrectly verifies, the cryptographic signature for data.

Source: MITRE CWE corpus.

CVSS Vector

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N

Timeline

Published
October 1, 2026
Last Modified
October 1, 2026
First Seen
October 1, 2026

Related Vulnerabilities