Test runs

A test run is a single execution against a project. This guide covers the most common kind — an autonomous exploratory run — and how to read its results. (Running specific test cases or a plan is covered in Test cases and Test plans.)

What it is#

During an exploratory run, BugBrain's AI agents open a real browser and explore your app like a careful user: they map the reachable pages, choose high-value flows to exercise, drive them end-to-end, and watch for failures — broken interactions, console errors, failed network calls, dead ends, accessibility problems, and more. Everything is captured with screenshots and a step-by-step log.

Why use it#

• Coverage without authoring — find problems across your whole app before you've written a single test case.
• Real-user behavior — the agent interacts the way a person would, surfacing issues scripted tests miss.
• A baseline — the first runs map your app and reveal the obvious bugs, giving you a foundation to build test cases on.

Before you start#

You need a project with a reachable URL. To test signed-in areas, add login credentials in the project (see Projects). You need the test-runs:create permission to start runs.

Start a run#

1. 1

   Open the project

   Go to the project's Overview or Test Runs page.
2. 2

   Start a new run

   Choose Start a run. For an exploratory run, that's all you need — the agent decides what to test.
3. 3

   Watch live progress

   The run appears under Active Runs with live status: pages visited, flows completed, and issues found so far.

Read the run viewer#

Open any run — live or finished — to see everything it did.

• Status — RUNNING, COMPLETED, FAILED, or CANCELED.
• Timeline & screenshots — each page visited and action taken, with a snapshot and an overlay of what the agent clicked.
• Issues found — the problems surfaced, each with severity, category, and evidence. Click through to the issue detail.
• AI action log — the agent's reasoning step by step (what it thought, what it did, what it observed), so a result is never a black box.
• Artifacts — console and network logs, and a session recording when available.

Manage runs#

• History — the Test Runs page lists past runs, filterable by outcome (passed / failed). Open any to inspect it.
• Cancel — stop an in-progress run from its viewer or the active-runs list.
• Rerun — start a fresh run anytime; results are versioned, so trends build up over time.
• Triggers — runs can be started manually, on a schedule, or by a pull request. The trigger is shown on each run.

Tips#

• Keep the project's start URL pointed at a stable, representative environment.
• Use project settings to bound exploration (max flows, depth) so runs stay fast and focused.
• Once exploratory runs reveal your real flows, promote the important ones to test cases for repeatable coverage.

Related#