Preview. This part of the console is still being built. What you see here describes the intended behavior and can change before general availability.
Tests Center
Scanning a repository is harmless. Attacking a running API is not, and nobody wants to learn that a test ran against production by reading the incident report.
So Tests Center puts a person in the path. Each test moves through fixed stages, and nothing reaches your application before an approver says so.
The stages
| Stage | What happens |
|---|---|
| Context | You describe the application and what the test should care about. |
| Recon | Dwaar learns the target. Recon belongs to the resource, so a later test reuses its profile. |
| Plan | Dwaar proposes what to run. |
| Approval | An approver reviews the plan. |
| Run | The Outpost carries out the run. |
| Findings | Results arrive as findings on the resource. |
The engines
| Engine | Tests | What a reviewer approves |
|---|---|---|
| ZAP | APIs | A scripted plan file |
| Strix | APIs | AI led testing under rules of engagement |
| Dforge | AI agents | A test suite |
The tabs
- Tests lists every test, grouped by who it is waiting on.
- Approvals holds the recon and test plans waiting for an approver.
- Runs shows every recon and test run the Outpost has done, is doing or will do.
- Catalogue lists attack vectors: the categories of checks each engine can run. A plan selects from these.
- Settings decides who approves, when tests may run, and the limits every run keeps.
Try this next
Set your approvers and allowed test windows in Settings before anyone creates a first test. Doing it afterwards means having that conversation in a hurry.