CVE-2026-111321: TaskingAI through 0.3.0 SSRF via OpenAPI servers[0].url in Actions
TaskingAI through 0.3.0 contains a server-side request forgery vulnerability that allows authenticated attackers to make the backend send arbitrary HTTP requests by supplying an unvalidated OpenAPI servers[0].url in a user-defined action. Attackers holding a project API key or console admin login can create an action pointing to loopback, internal or cloud metadata addresses and run it via POST /actions/{actionid}/run to read back up to 64 KiB of responses.
Affected Software
Event History
Frequently Asked Questions
Who can exploit this issue?
An attacker needs valid TaskingAI credentials: either a project API key or a console administrator login. The issue is therefore relevant where those credentials are available to untrusted or lower-trust users.
What access can an attacker reach through the vulnerable request?
An attacker can configure an action to send HTTP requests to loopback addresses, internal network services, or cloud metadata endpoints. Responses can be returned through the action execution endpoint, with up to 64 KiB readable per response.
How is the vulnerable behavior triggered?
The attacker creates a user-defined action whose OpenAPI servers[0].url points to the target address, then runs that action using POST /actions/{action_id}/run. Exploitation does not require user interaction.