← ClaudeAtlas

e2e-testslisted

Generate automated API/E2E tests for features that already exist — brownfield code, features shipped before the eval net, or manual test-plan scenarios worth automating. Use when the user says "generate e2e tests", "backfill tests", "automate the test plan", or "qa tests for {feature}".
rterakedis/leanwheel-skills · ★ 2 · Testing & QA · score 65
Install: claude install-skill rterakedis/leanwheel-skills
# E2E Test Generation Skill **Goal:** Retro-fit an automated regression net onto already-built features. All story-scoped testing in this workflow is forward-looking — `evals` BUILD derives cases from ACs *as stories are written*. This skill covers everything that shipped without that coverage: brownfield projects onboarded via `/discover`, features built before `docs/evals/` existed, and manual test-plan scenarios that a machine could run instead of a human. **Not this skill:** code review (`/code-review`), story-scoped test authoring (`/dev-story` writes those), visual verification (`/design-verify`). **Token posture:** one authoring session; every generated test is registered as a `type: command` eval case — zero model tokens on every future RUN, forever. --- ## Step 1 — Detect the test framework Use what the project already has — never introduce a parallel framework: - **Swift/Apple:** existing XCTest / Swift Testing targets (`Package.swift` test targets, `.xctest` schemes); UI flows use **XCUITest**. Read `docs/setup/swift/testing.md` if present for project conventions. - **Web:** `package.json` dev-dependencies (playwright, cypress, vitest, jest); existing `tests/` or `e2e/` dirs for patterns. - **Python:** pytest config; **API-only:** `curl`-based checks are acceptable. If no framework exists, propose the stack standard (XCTest/XCUITest for Swift, Playwright for web UI, pytest for Python, vitest+supertest for Node APIs) and **confirm with the user before