Summary
MindsDB through version 26.1.0 contains a server-side request forgery vulnerability in the web crawler handler. An unauthenticated attacker may supply a caller-controlled URL to CrawlerTable.list and cause the server to request internal services or cloud metadata endpoints when the allowlist is empty or insufficiently enforced.
Affected context
The vulnerability occurs in MindsDB deployments that expose the web crawler functionality and permit unauthenticated requests to reach the CrawlerTable.list handler. Risk is increased when the crawler accepts arbitrary destination URLs and the default or active allowlist does not restrict outbound requests to approved external destinations.
Exploit preconditions
- Attack surface: Web-accessible MindsDB crawler functionality
- Required access: No authentication may be required when the affected handler is exposed
- Trigger point: A request to CrawlerTable.list containing a caller-controlled URL
- Why exploitation is possible: The server makes outbound requests based on attacker-supplied destinations, while an empty or ineffective allowlist fails to constrain internal and metadata-oriented targets
Why this happens
The underlying bug is server-side request forgery, mapped to CWE-918. The application treats a user-provided URL as a permitted outbound destination and performs the request from a trusted server-side network location. When destination validation is absent, incomplete, or permissive by default, attackers can use the server as a proxy to reach resources that are not directly accessible from the public network.
Potential impact
- Internal service responses may be exposed to an attacker
- Cloud instance metadata or temporary credentials could be disclosed
- Internal network topology and service availability may be enumerated
- Sensitive administrative or application endpoints could be accessed
- Downstream systems may be affected by requests originating from a trusted network location
General mitigation guidance
Mitigation typically focuses on reducing trust in caller-supplied destinations and enforcing outbound network boundaries.
Common defensive patterns include:
- Applying strict allowlists for schemes, hostnames, ports, and resolved addresses
- Rejecting loopback, link-local, private, reserved, and metadata-service destinations
- Revalidating destinations after DNS resolution and across redirects
- Requiring authentication and authorization for crawler functionality
- Restricting egress from application workers through network policy or firewall controls
- Using secure-by-default configurations that deny outbound requests until destinations are explicitly approved
- Running crawler processes with minimal network and operating-system privileges
How teams detect issues like this
Organizations commonly detect SSRF risks through:
- Static analysis rules identifying HTTP clients that receive URLs from request parameters
- Code review checks for missing scheme, hostname, port, redirect, and IP-range validation
- Tests covering private, loopback, link-local, metadata, encoded, and redirecting destinations
- Runtime monitoring of outbound requests to internal address ranges and cloud metadata endpoints
- Policy enforcement that blocks unrestricted egress from application workloads
- Automated security review tools that flag user-controlled URLs passed to network request APIs

