skills.rest
Assamir
AssamirCOMMUNITY
Explore all skills by this author →
GitHub

10x-e2e

Generates and verifies risk-driven Playwright E2E tests against a running application.

Updated Jul 4, 2026
Search Tags:#test generation#browser automation#playwright#e2e testing#test review#ci#risk-driven
One-click install
npx skills add https://github.com/Assamir/ai-toolkit --skill 10x-e2e-assamir
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: 10x-e2e
Source: https://github.com/Assamir/ai-toolkit/tree/main/.cursor/skills/10x-e2e
Command: npx skills add https://github.com/Assamir/ai-toolkit --skill 10x-e2e-assamir

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Agent-generated E2E tests often pass without protecting anything: they use brittle CSS selectors, hardcoded waits, shared state, and assertions that never fail when the real risk materializes. This Skill drives an approved plan's browser-level phases one risk at a time through a plan → generate → review → verify loop, so each Playwright test is tied to a named risk and proven to fail when that risk occurs. ## Core Features & Use Cases - Risk-driven E2E loop: For each browser-level risk in a plan, it maps the flow against the running app, generates a test from a seed exemplar and E2E rules, reviews it against five agent anti-patterns, and verifies it with a deliberate-break check. - Eligibility gating and redirects: Before each phase it checks browser-level fit, feature presence, and test absence, redirecting non-E2E work to /10x-tdd or /10x-implement and sharing the same plan Progress state with those skills. - Quality levers and references: Ships a seed test pattern, E2E quality rules, an anti-pattern review checklist, a prompt template, and browser-driven generation guidance (Playwright CLI or MCP, accessibility-tree based). - Use Case: After building a feature with /10x-implement, run /10x-e2e save-session phase 6 to produce one reviewed, break-verified Playwright spec proving that saved data survives a full page reload, then commit it with the phase-end ritual. ## Quick Start Ask the assistant to run /10x-e2e with your change-id and phase number, for example to drive the E2E phase of an approved plan in context/changes.

Frequently Asked Questions about 10x-e2e

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate Playwright E2E tests from a test plan?▼

Run /10x-e2e with a change-id or plan path, and it reads the plan's Progress section and test-plan.md risks, then loops plan → generate → review → verify per browser-level risk. Each test is modeled on a seed spec and verified with a deliberate-break check before being marked done.

When should I use E2E tests instead of unit tests?▼

Use E2E when a risk crosses several system boundaries like auth, routing, API, and database, or exists only in the rendered UI. If an isolated function or integration test can prove the risk, the skill redirects that phase to /10x-tdd or /10x-implement instead.

Does this skill set up Playwright for my project?▼

No, it assumes Playwright is already installed with a config, a single-spec run command, storageState auth, and a way to start the app. If no Playwright config or spec files exist, it stops and tells you to install Playwright first.

Can I use this with Cypress or WebdriverIO instead of Playwright?▼

Yes, the principles transfer to other stacks. You encode your tool's idioms, such as its getByRole equivalent and wait-for-state mechanism, into your own seed test and rules file, then drive that tool's runner.

Why do agent-generated E2E tests become flaky?▼

They typically exhibit five anti-patterns: hallucinated assertions, brittle CSS selectors, shared state between tests, waitForTimeout instead of state-based waits, and missing cleanup. The skill reviews every generated test against these and re-prompts by naming the specific anti-pattern.

What is the deliberate-break verification in E2E testing?▼

After a test passes, you temporarily invert the production behavior the risk targets and re-run the test to confirm it goes red. If it stays green, the assertion protects nothing and must be fixed; the break is reverted immediately and never committed.

Related Skills

Explore More Software Engineering Skills→
BMS-kmusialBMS-kmusial

10x-goal-implement

Implements technical plans autonomously with per-phase quality gates and Conventional Commits.

Community
Advanced
wishx127wishx127

openspec-archive-change

Archive completed OpenSpec changes after verifying artifacts and task status.

Community
Advanced
samcdavidsamcdavid
4

my-implement

Execute approved implementation plans with red/green/refactor TDD validation.

Community
Advanced
distribucionesfayosmoreno-sysdistribucionesfayosmoreno-sys

ui-visual-validator

Validate UI modifications for visual regression, compliance, and accessibility.

Community
Advanced