Skip to main content
Vulnerability Database/CVE-2026-114089

CVE-2026-114089: Trigger.dev CLI CSRF Vulnerability

CVE-2026-114089 is a CSRF flaw in Trigger.dev CLI that allows attackers to steal Personal Access Tokens through malicious authorization-code URLs. This post explains its technical details, affected versions, and mitigation steps.

Published:

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

CVE-2026-114089 Overview

CVE-2026-114089 is a cross-site request forgery (CSRF) vulnerability affecting Trigger.dev versions before 4.5.2. The flaw resides in the command-line interface (CLI) authorization-code flow. Loading /account/authorization-code/<code> mints a Personal Access Token (PAT) without any consent step. Attackers mint an unauthenticated code, lure a logged-in user to that URL, and then retrieve the victim's PAT from the unauthenticated /api/v1/token endpoint. The resulting token grants full account takeover. The weakness is classified under CWE-352.

Critical Impact

An attacker who lures an authenticated Trigger.dev user to a crafted URL obtains a Personal Access Token and takes over the victim's account.

Affected Products

  • Trigger.dev versions prior to 4.5.2
  • Trigger.dev CLI authorization-code flow
  • Trigger.dev /api/v1/token endpoint

Discovery Timeline

  • 2026-10-21 - CVE-2026-114089 published to NVD
  • 2026-10-21 - Last updated in NVD database

Technical Details for CVE-2026-114089

Vulnerability Analysis

Trigger.dev implements a CLI authorization-code flow that lets developers link local CLI sessions to their accounts. The flow mints a Personal Access Token when a logged-in user loads /account/authorization-code/<code>. The server performs no consent check and does not validate the origin of the request. An attacker generates an authorization code from an unauthenticated context, then tricks a logged-in victim into visiting the crafted URL. The victim's browser carries session cookies, so the backend binds a PAT to the attacker-controlled code. The attacker then polls the unauthenticated /api/v1/token endpoint with the same code to retrieve the newly minted token. The PAT grants programmatic access to the victim's workloads, triggers, environment variables, and project data.

Root Cause

The root cause is the absence of a user-facing consent step and anti-CSRF controls in the authorization-code endpoint. The endpoint treats any authenticated GET request as implicit approval to mint and bind a token to a caller-supplied code. The companion /api/v1/token endpoint compounds the issue by returning the bound token without authenticating the caller.

Attack Vector

Exploitation requires user interaction over the network. An attacker hosts a page or sends a link that references /account/authorization-code/<code> using an image tag, iframe, redirect, or phishing lure. When the authenticated victim loads the URL, the server mints a PAT. The attacker then issues an unauthenticated request to /api/v1/token with the same code and receives the token. Refer to the GitHub Security Advisory GHSA-58mc-m3qq-6qrh and the VulnCheck Advisory for additional analysis.

Detection Methods for CVE-2026-114089

Indicators of Compromise

  • Unexpected GET requests to /account/authorization-code/<code> originating from external referrers or embedded in third-party pages.
  • Unauthenticated requests to /api/v1/token that successfully return a PAT shortly after a user session loads an authorization-code URL.
  • New Personal Access Tokens appearing in user account token lists without a corresponding CLI login event initiated by the user.
  • API activity from unfamiliar source IP addresses or user agents immediately after a PAT is minted.

Detection Strategies

  • Correlate web access logs for /account/authorization-code/ with Referer headers that are external to the Trigger.dev domain.
  • Alert on token minting events that are not preceded by a legitimate CLI-initiated code generation workflow.
  • Baseline PAT creation volume per user and flag deviations, particularly tokens issued and consumed within short time windows.

Monitoring Recommendations

  • Forward Trigger.dev application and reverse-proxy logs to a central analytics platform and retain them for retrospective review.
  • Monitor outbound API usage from newly minted PATs and alert on first-use source IPs that differ from historical developer locations.
  • Track administrative actions such as environment variable changes, deployment triggers, and project membership edits performed by tokens created within the last 24 hours.

How to Mitigate CVE-2026-114089

Immediate Actions Required

  • Upgrade Trigger.dev to version 4.5.2 or later across all self-hosted and managed deployments.
  • Revoke all existing Personal Access Tokens and require users to regenerate tokens through the patched flow.
  • Audit account activity, triggers, and environment variables for unauthorized changes dating back to the first exposure of the vulnerable endpoint.
  • Notify users to review their active sessions and connected CLI devices.

Patch Information

The fix is available in Trigger.dev version 4.5.2. The patched release adds a required consent step to the CLI authorization-code flow and removes the ability to retrieve a bound token from an unauthenticated endpoint. Refer to the GitHub Security Advisory GHSA-58mc-m3qq-6qrh for upgrade guidance.

Workarounds

  • Restrict access to /account/authorization-code/ and /api/v1/token at the reverse proxy or web application firewall until the patch is applied.
  • Enforce strict SameSite=Strict cookies on session cookies to reduce cross-site request feasibility.
  • Educate developers to avoid clicking authorization-code links from untrusted sources while logged into Trigger.dev.
bash
# Configuration example
# Example NGINX rule to block external access to the vulnerable endpoints
location ~ ^/account/authorization-code/ {
    if ($http_referer !~* "^https://your-trigger-domain\.example\.com/") {
        return 403;
    }
}

location = /api/v1/token {
    deny all;
}

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

Experience the Most Advanced Cybersecurity Platform

See how the world’s most intelligent, autonomous cybersecurity platform can protect your organization today and into the future.