Summary
A cross-site request forgery vulnerability in Budibase's chat-link handoff endpoint allows a crafted web page to submit an unauthorized request in a victim's browser. If a confirmation token is available to the attacker, the request may bind the attacker's external chat identity to the victim's account, enabling impersonation within agent operations and inheritance of the victim's permissions.
Affected context
The issue occurs in Budibase deployments using the chat-link handoff workflow. Risk is present when the endpoint accepts browser-originated state-changing POST requests without sufficiently enforcing request origin or anti-CSRF protections, and when a confirmation token can be obtained or exposed to an attacker.
Exploit preconditions
- Attack surface: The chat-link handoff HTTP endpoint
- Required access: Ability to deliver a phishing page or otherwise cause a logged-in victim's browser to submit a request; access to a valid confirmation token may also be required
- Trigger point: A state-changing POST request that binds an external chat identity to an account
- Why exploitation is possible: The endpoint does not adequately distinguish an intentional user action from a cross-site request made using the victim's authenticated browser context
Why this happens
The vulnerability is caused by insufficient protection of a state-changing web request against cross-site submission. When an endpoint relies primarily on ambient browser credentials and does not robustly validate request intent through CSRF tokens, origin checks, or equivalent controls, another site may cause the browser to perform an action on the user's behalf. This behavior maps to CWE-352: Cross-Site Request Forgery.
Potential impact
- An attacker may link their external chat identity to a victim's account
- An attacker could impersonate the victim during agent operations
- The attacker may inherit permissions associated with the victim's account
- Unauthorized actions could be performed using the victim's authenticated context
- Account relationships or audit records could be altered without the victim's intent
General mitigation guidance
Mitigation generally focuses on ensuring that sensitive account-linking actions require demonstrable user intent and cannot be initiated solely through ambient browser credentials.
Common defensive patterns include:
- Applying strong, single-use CSRF tokens to all state-changing requests
- Validating the request Origin or Referer where appropriate as a defense-in-depth control
- Using restrictive cookie SameSite settings consistent with application requirements
- Requiring explicit user confirmation for identity-binding operations
- Keeping confirmation tokens short-lived, scoped, non-reusable, and protected from exposure
- Applying least privilege to linked identities and requiring reauthentication for high-impact account changes
- Avoiding state-changing actions through cross-site-compatible request flows
How teams detect issues like this
Organizations can identify similar issues through:
- Static analysis and code review for state-changing endpoints lacking CSRF validation
- Review of authentication and account-linking flows for reliance on ambient cookies
- Testing cross-origin submissions with and without valid CSRF tokens
- Monitoring for unusual identity-linking events, unexpected origins, and rapid changes to account associations
- Policy enforcement requiring anti-CSRF controls for browser-accessible state-changing routes
- Automated security testing or code review tools that flag missing request-intent validation

