In Dradis Community Edition the authorization check for creating AI provider configurations is omitted because a constant used in the before_action is undefined in the community build. This oversight lets any authenticated, non‑administrator user add a provider that points to an arbitrary HTTP or HTTPS address, then reassign the built‑in Roslin agent to use that provider. When the agent is invoked, the server performs a request to the supplied URL. If the response is a non‑2xx code, the body is returned to the attacker through ActionCable/Turbo Stream error messages, making the internal response payload discoverable. The vulnerability is therefore a Server‑Side Request Forgery that can reveal internal services or other information that would normally be inaccessible from the user’s browser.

The flaw exists in Dradis Community Edition versions 5.1.0 through 5.2.0. No other product variants are listed. The affected package is the community edition of Dradis, identified by the vendor product name dradis:dradis‑ce. No specific application sub‑versions beyond the version range are required.

The CVSS score is 7.1, indicating a high severity. EPSS information is not currently available, so the likelihood of exploitation cannot be quantified, but the vulnerability is not listed in the CISA Known Exploited Vulnerabilities catalog, implying no publicly demonstrated exploits at the time. Attackers would need an authenticated account within the application to create a provider. Once set, the SSRF can be triggered by any user interaction that causes the Roslin agent to query the provider, leading to potential disclosure of internal network information or other services. Because the vulnerability arises from missing authorization rather than a direct code execution path, exploitation requires only the creation of a privileged look‑ups rather than malicious input delivered via the web interface.