What problem does it solve? Manually writing and maintaining end-to-end tests is slow, and coding agents lack a structured way to verify that the code they produce actually works against a live application. This Skill drives the TestSprite CLI to create, run, and verify real browser and API tests, closing the loop between code changes and verified behavior. ## Core Features & Use Cases - Test Creation & Execution: Create frontend (Playwright-based) and backend API tests from plan files, run them in the cloud, and wait for structured JSON results with clear exit codes. - Failure Diagnosis Loop: Download self-contained failure bundles (failure.json, screenshots, DOM snapshots, test source) with root-cause hypotheses, then rerun after fixes. - CI & Agent Integration: Non-interactive authentication, dependency-aware backend test waves, batch operations, and JSON output for scripting in CI pipelines. - Use Case: An agent implements a new checkout feature, creates a frontend test from a plan JSON, runs it, receives a failure bundle showing a changed button selector, fixes the code, and reruns until the test passes. ## Quick Start Install the TestSprite CLI with npm, run testsprite init with your API key, then ask the agent to create and run a frontend test for your checkout flow and report any failures.
