Skip to main content
Version: 1.0.7

Quick Start

This guide follows the shortest safe path from an empty workspace to execution evidence.

Before you begin​

Confirm that you have:

  • access to the intended RabbitQA workspace;
  • permission to create or use an AutoRunner source and Test Plan;
  • a target environment that is safe for automated actions;
  • non-production test credentials and data, when the flow needs them; and
  • an available runner or project engine configuration.

1. Connect a source​

Open AutoRunner > Sources > Connect Source. Choose the project type, framework, and either a RabbitQA template or your configured repository.

AutoRunner source configuration

Project type and framework define how the source is handled. Verify both before creation because they are not intended to be changed later. For AI Studio work, do not create an AI_AGENT source manually; AutoRunner manages that source behind AI Studio.

2. Prepare a scenario​

Use either path:

  • Scenario List for scenarios synchronized from connected sources; or
  • Scenario List > AI Studio for AI-assisted Web or Mobile scenarios.

In AI Studio, select New Scenario, give it a unique name, choose a platform, and write a bounded prompt that states what the agent may and may not do.

Bounded AI scenario prompt

For Web generation, confirm the target URL before selecting Generate.

Web generation setup

Review the generated steps, warnings, conversation, and version state. Correct review-required steps before publishing. A passing generation alone is not execution proof.

3. Configure test data​

Open Test Datas and add only the data needed by the scenario.

Variable precedence is:

  1. runtime override;
  2. selected environment value; and
  3. base variable.

Use masked visibility for secrets. You can also upload supported files or import approved files and variables from DataCrate.

4. Create a Test Plan​

Open Test Plans > New Test Plan. Add at least one scenario, order the selected scenarios, and optionally select a default Run Profile or an environment-profile override. A plan can select scenarios across sources in the active workspace.

5. Run and verify​

Run the plan from Test Plans or with an appropriate Run Profile. Open Test Runs to track the execution.

Test Run filters and actions

Verify:

  • final status and result;
  • scenario and step outcomes;
  • logs and timestamps;
  • screenshots or artifacts when produced; and
  • the runner, environment, trigger, and profile used.
Do not test with unreviewed production data

AI generation, scenario execution, schedules, issue creation, and source operations can affect external systems. Use an approved target, bounded credentials, and reversible test data.

Continue with Test Plans, Test Runs, and Run Detail.