Join our Discord
Run workflows with Ask AI
Drive multi-step product workflows from a single natural-language instruction using Ask AI on the BrowserStack MCP server.
Use Ask AI to hand over an entire job rather than a single action. Give one instruction from your AI assistant, such as GitHub Copilot, Claude, or Cursor, and Ask AI works out which BrowserStack capabilities are needed, runs them in sequence, and reports back with a link to the result.
The individual tools on this MCP server each perform one deterministic action. Ask AI is for requests that span several of them, or that you can’t map to specific tools in advance.
To set up the server, see get started with the BrowserStack MCP server.
To use Ask AI on the MCP server, you need a Test Management or Test Reporting & Analytics license, depending on the workflow. Contact BrowserStack support to enable a license for your account.
What Ask AI can do
Ask AI supports Test Management and Test Reporting and Analytics workflows.
Test Management workflows
Ask AI handles the following Test Management workflows:
- Search and filter test cases in natural language.
- Create and manage test cases, runs, plans, folders, and tags.
- Generate reports and dashboards, and interpret what the data means.
- Review how a test case changed over time.
- Enrich or clean up poorly written test cases.
Try the following example prompts:
- Add the tag ‘regression’ to all test cases in the Payments folder and list any that have never been run.
- Find test cases related to payment processing.
- How has this test case evolved over the last 3 versions?
Test Reporting & Analytics workflows
Ask AI handles the following Test Reporting & Analytics workflows:
- Run root cause analysis on a build or an individual test.
- Assess release readiness, with genuine failures separated from flaky tests and environment noise.
- Find and fix flaky and consistently failing tests, with suggested code changes.
- Build and edit dashboards.
- Apply filters and save reusable views.
- Ask open questions about test quality and trends.
Try the following example prompts:
- Debug this build, group the failures by theme, and tell me whether it’s safe to ship.
- Why did this test fail, and how do I fix it?
- Which tests waste the most CI time?
Ask AI plans and runs multiple steps from a single instruction, and it doesn’t ask for confirmation before each step. Keep the following in mind before you rely on its output:
- It can apply bulk changes in a single pass, such as tagging many test cases at once. Review the actions and the linked records Ask AI returns, especially before bulk or hard-to-reverse updates.
- For a release-readiness call or a go or no-go call, treat the recommendation as a starting point. Verify it against the build evidence before you act on it.
Tool reference
Use these example prompts to interact with Ask AI from your AI assistant. Every prompt calls the same askBrowserStackAI tool, with different input parameters depending on the workflow. You don’t set these parameters yourself. Your AI assistant fills in prompt and either product or context, depending on the workflow, when it calls the tool on your behalf. You can verify the tool call in the MCP server logs.
| Example prompt | Description | Input parameters | Invoked tool |
|---|---|---|---|
| Find all payment test cases in the ‘Shopping App’ project and add the ‘regression’ tag to them | Runs a multi-step Test Management workflow from a single instruction. Ask AI resolves the entities referenced in the prompt, plans the required steps, and executes them in order. Ask AI returns a summary of what it did, and a link to the affected records in Test Management. |
• prompt (required): The instruction in natural language• product (required): test-management |
askBrowserStackAI |
| Debug my latest build, separate genuine failures from flaky ones, and give me a go or no-go call | Runs a multi-step Test Reporting & Analytics workflow. Ask AI pulls the run evidence, performs root cause analysis, classifies failures, and returns a recommendation. Ask AI returns the root cause of each failure, a recommended fix where available, and a release-readiness call. |
• prompt (required): The instruction in natural language• product (required): test-reporting-analytics |
askBrowserStackAI |
| Build me a dashboard showing flaky percentage by team for the last 30 days | Creates or edits a dashboard from a description. Ask AI selects the widgets and filters, names the dashboard, and returns a link to it. Follow-up prompts refine the same dashboard. | • prompt (required): The dashboard description• product (required): test-reporting-analytics |
askBrowserStackAI |
| Summarize how the ‘Invalid Login Scenario’ test case has changed over its last three versions | Answers questions about your Test Management or Test Reporting & Analytics data without changing anything. Read-only prompts return an answer grounded in your account data. | • prompt (required): The question in natural language• context (required): URL or ID of the entity your question refers to |
askBrowserStackAI |
Related topics
For more help, see the following pages:
We're sorry to hear that. Please share your feedback so we can do better
Contact our Support team for immediate help while we work on improving our docs.
We're continuously improving our docs. We'd love to know what you liked
We're sorry to hear that. Please share your feedback so we can do better
Contact our Support team for immediate help while we work on improving our docs.
We're continuously improving our docs. We'd love to know what you liked
Thank you for your valuable feedback!