Description

Budibase Server before 3.41.3 contains a server-side request forgery vulnerability in the datasource verify endpoint that allows builder-level users to supply arbitrary URLs without SSRF validation. Attackers can exploit this to leak internal CouchDB credentials by making requests to attacker-controlled servers, gaining full database access in cloud deployments.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "7.6",
  "pubDate": "2026-08-28T12:16:34.477Z",
  "pubdate": "2026-08-28T12:16:34.477Z",
  "executiveSummary": "Budibase Server versions prior to 3.41.3 contain a critical Server-Side Request Forgery (SSRF) vulnerability located within the datasource verify endpoint. This flaw permits authenticated users with builder-level privileges to bypass security controls by supplying arbitrary URLs that lack proper validation or sanitization. By directing the server to interact with external, attacker-controlled infrastructure, an adversary can coerce the application into leaking sensitive configuration data, specifically internal CouchDB credentials. The successful exploitation of this vulnerability poses a severe risk to cloud deployments, as the acquisition of these credentials grants an attacker full, unauthorized administrative access to the underlying database system. This compromise allows for the exfiltration, modification, or destruction of stored data, significantly undermining the confidentiality, integrity, and availability of the Budibase environment. The vulnerability highlights a failure in input validation when handling server-initiated network requests, enabling attackers to leverage the server as a proxy to reach internal resources or malicious endpoints.",
  "technicalDetails": "The vulnerability originates from a deficiency in input validation within the Budibase Server datasource verify endpoint. When a user initiates a datasource configuration verification, the application backend performs an outbound HTTP request to the provided URL to confirm connectivity or schema validity. In versions prior to 3.41.3, this implementation fails to enforce an allow-list of permitted domains or validate the destination against internal network ranges, allowing for unconstrained SSRF.\nExploitation requires an authenticated session with at least builder-level privileges. The attacker triggers the vulnerability by manipulating the datasource verification request, injecting a URL pointing to an attacker-controlled listener or a proxy service. When the Budibase server processes the verification request, it initiates an outbound request to the attacker-supplied URL.\nThe attack flow proceeds as follows: 1) The attacker authenticates to the Budibase dashboard. 2) The attacker navigates to the datasource configuration interface. 3) The attacker initiates a connection verification, intercepting the request or using the UI to provide a crafted URL pointing to an attacker-controlled server. 4) The server-side component retrieves the target URL without inspecting the destination for prohibited protocols or internal hostnames. 5) The server potentially reflects sensitive headers or authentication tokens, or in specific configurations, leaks internal CouchDB credentials as part of the request payload or subsequent server interaction. 6) The attacker captures these credentials from their external listener.\nThis vulnerability is particularly impactful in cloud-native deployments where Budibase is tightly integrated with CouchDB. By obtaining the administrative credentials for CouchDB, the attacker gains complete control over the persistence layer, effectively bypassing all application-level authorization and access control logic. The ability to perform arbitrary requests from the server context also provides the attacker with a pivot point into the internal network, potentially facilitating further lateral movement or interaction with internal microservices that assume the server is a trusted entity.\nThe root cause is the reliance on user-supplied input for constructing backend network requests without implementing a robust server-side request filtering mechanism. By failing to restrict the outbound request to trusted endpoints, the application effectively acts as an open proxy for the attacker."
}