Autopilot best practices

Guidance for improving Autopilot accuracy in Test Manager when evaluating requirements, generating test cases, and automating manual tests.

Autopilot helps automate and enhance tasks across your testing lifecycle. Follow these best practices to improve accuracy and usability.

Note:

Autopilot is available only in Test Manager delivered via Test Cloud.

Requirement evaluation

For evaluating requirements you need to:

• Write clear, complete requirements with measurable acceptance criteria.
• Use focused evaluations (for example: security or performance).
• Attach supporting files like guidelines or specs.
• Use the Prompt Library to standardize evaluations.

Manual test generation

For generating manual tests you need to:

• Ensure requirements include full user flows and expected outcomes.
• Add supporting files (mockups, diagrams) for context.
• Tailor instructions for specific scenarios.
• Reuse prompts from the Prompt Library.

Converting text into code

For converting text into code you need to:

• Specify the language and goal clearly (e.g., “Refactor this C# method”).
• Keep prompts short and direct.

Automating manual tests

For automating manual tests you need to:

• Use a consistent object repository and naming conventions.
• Write manual steps using terms that match UiPath activity names.

Generating test data

For generating test data you need to:

• Use prompts to define value ranges, patterns, or combinations.
• Reference existing arguments in your workflows.

Analyzing test results

For analyzing test results you need to:

• Run insights on a large volume of test results for better accuracy.
• Focus on:
  • Common Errors: Frequent failure clusters.
  • Error Patterns: Recurring test failure themes.
  • Recommendations: Actionable fixes to stabilize tests.