Summary
Baserow's Application Builder data-source endpoints can proceed without enforcing the result of a permission check. An unauthenticated caller may therefore request a data source and receive rows and fields retrieved using the integration's own credentials when the relevant identifiers are discoverable.
Affected context
The issue occurs in Baserow Application Builder data-source dispatch and record-name views when endpoint permissions admit unauthenticated requests and the service does not act on denied permission results. Risk is present where data-source identifiers can be obtained or enumerated and the associated integration can read sensitive records or fields.
Exploit preconditions
- Attack surface: Application Builder data-source dispatch and record-name API endpoints
- Required access: No valid application credential may be required by the affected endpoint
- Trigger point: A request that identifies a data source and causes it to be dispatched
- Why exploitation is possible: The permission-check helper returns denials without raising, while the caller neither stores nor evaluates the denial mapping, allowing execution to continue under the integration's credentials
Why this happens
The underlying bug is an authorization-enforcement failure mapped to CWE-862: Missing Authorization. Although the service invokes a permission-checking routine, it does not request exception-based enforcement or inspect the returned denials. As a result, a failed authorization decision has no effect on control flow, and the data-source operation proceeds using trusted integration credentials rather than the requester's identity.
Potential impact
- An unauthenticated user may access rows and fields exposed by a data source
- Sensitive business information could be disclosed
- Data-source boundaries may be bypassed across Application Builder resources
- Exposed information may support further account, tenant, or system compromise
General mitigation guidance
Mitigation generally focuses on making authorization checks fail closed and ensuring every denial is enforced before data access occurs.
Common defensive patterns include:
- Require authentication and appropriate resource permissions at the endpoint boundary
- Use authorization APIs that raise or return an outcome that is explicitly checked
- Centralize authorization enforcement so handlers cannot accidentally ignore denial results
- Bind data access decisions to the caller's identity and tenant context rather than relying only on integration credentials
- Apply least privilege to integrations and limit the fields and records each data source can retrieve
- Use non-guessable resource references as a defense-in-depth measure, without treating obscurity as authorization
How teams detect issues like this
Organizations commonly identify missing-authorization risks through:
- Static analysis for permission checks whose return values are ignored
- Code review searches for handlers that accept anonymous callers while dispatching protected resources
- Tests that exercise unauthenticated, cross-user, and cross-tenant requests against every data-source operation
- Policy enforcement requiring explicit authorization decisions before service-layer data access
- Monitoring for repeated requests across sequential or otherwise enumerable data-source identifiers
Automated application-security and code-review tools can help flag ignored authorization results, but runtime authorization tests remain important.

