Quickstart

Documentation Index

Fetch the complete documentation index at: https://docs.testsprite.com/llms.txt

Use this file to discover all available pages before exploring further.

By the end of this guide you’ll have a UI project that’s been explored, planned, generated as Playwright code, and run against your live app, with a recorded video for every test. No code on your side.

​

Step 1: Project Setup & PRD Upload

From the dashboard, click Create Tests. The first screen asks for two things only: a project name and your PRD (or any product-spec document — markdown, PDF, plain text all work).

​

Step 2: Feature & Use-Case Extraction

If you uploaded a PRD, TestSprite analyzes it and shows the Use Cases & Features screen with a flow graph of every feature it identified and the use cases under each. Stages while it’s working: Reading your documentation → Identifying key features → Mapping use cases. A toast confirms the count when it’s done (e.g. “Found 12 features and 47 use cases”). This feature map is what grounds everything downstream — exploration targets these use cases, plan generation scopes test cases to them, and the report attributes coverage back to them. Click Re-extract to redo the analysis on the same PRD (e.g. after editing the file). Otherwise continue.

​

Step 3: Choose Test Type and Configure

Pick the test type tab — Frontend (URLs) for this UI walkthrough — then provide your starting URL, test account, and any extra testing instructions.

​

Step 4: Wait for Feature Exploration (Beta)

TestSprite walks your live app feature by feature, targeting the use cases from your feature map. You watch a live preview as it runs. Per-feature status: Exploring while it’s running, then Done, Failed, or Blocked when it finishes. You can retry individual features or update credentials without re-running everything. When you’re satisfied, click Continue to test plan.

​

Step 5: Review and Adjust the Plan

TestSprite drafts a customized test plan grounded in what was observed. Each row has a clear description explaining its purpose. Use Select All to toggle the whole batch. Untick anything irrelevant, edit descriptions in natural language to refine focus, then click Generate Tests.

​

Step 6: Watch Tests Generate and Run

TestSprite converts the approved plan into Playwright code per test case, then executes each test in our cloud sandbox. Tests stream into the project view as they finish. Each row surfaces one of: Pending, Running, Pass, Failed, Blocked.

​

Step 7: Replay Step-by-Step

Click any test row to open its detail page. The right pane plays the recorded video of the run; the Steps panel inside the Overview tab lists every action with a clickable HTML snapshot per step. For failures, the Error / Trace / Fix tabs surface the Playwright error, raw stack trace, and an AI-authored cause-and-fix suggestion.

​

Step 8: Refine and Rerun

To iterate on a test, open its detail page and use the Chat tab — describe what to fix in natural language. TestSprite either tweaks the assertions in place or regenerates the test. To re-execute, click Run (or Save & Run if you have unsaved edits) and confirm in the Rerun this test dialog. Pro plans get the Auto-heal on UI drift toggle there, which absorbs UI changes that don’t actually break the flow.

​

The Final Report

Once the run finishes, the Test Report tab summarizes pass / fail / blocked counts, highlights failures with cause-and-fix suggestions, and links into per-test detail. This is what you’ll share with the team.

​

Where to Go Next