The ultimate Cypress testing skill for AI coding agents.
Works with Claude Code, Codex CLI, OpenClaw, Cursor, and any agent that reads SKILL.md files.

What This Is

A production-grade AI agent skill for writing, debugging, and optimizing Cypress tests. Covers every major Cypress pattern — selector strategy, network stubbing, auth with cy.session, CI parallelization, component testing, flake elimination — in a format AI agents read and apply directly.

One SKILL.md at the root. No nested directories. Agents clone this repo and read SKILL.md immediately.

Quick Install

OpenClaw (global — available to all agents)

git clone https://github.com/YOUR_USERNAME/cypress-agent-skill \
  ~/.openclaw/skills/cypress-agent-skill

Claude Code (project-level)

git clone https://github.com/YOUR_USERNAME/cypress-agent-skill \
  .claude/skills/cypress-agent-skill

Claude Code (personal — available in all projects)

git clone https://github.com/YOUR_USERNAME/cypress-agent-skill \
  ~/.claude/skills/cypress-agent-skill

Codex CLI

git clone https://github.com/YOUR_USERNAME/cypress-agent-skill \
  .agents/skills/cypress-agent-skill

Cursor

git clone https://github.com/YOUR_USERNAME/cypress-agent-skill \
  .cursor/skills/cypress-agent-skill

Global (works with Codex, Gemini CLI, Kiro, Antigravity simultaneously)

git clone https://github.com/YOUR_USERNAME/cypress-agent-skill \
  ~/.agents/skills/cypress-agent-skill

One-liner bash installer

# Detects platform automatically
bash <(curl -fsSL https://raw.githubusercontent.com/YOUR_USERNAME/cypress-agent-skill/main/install.sh)

# Or with explicit agent
bash <(curl -fsSL https://raw.githubusercontent.com/YOUR_USERNAME/cypress-agent-skill/main/install.sh) --agent open-claw

LobeHub CLI

npx -y @lobehub/market-cli register --name "YourAgent" --source open-claw
npx -y @lobehub/market-cli skills install cypress-expert --agent open-claw

ClawHub (OpenClaw native)

clawhub install cypress-agent-skill

Repo Structure

cypress-agent-skill/           ← clone target / skill root
├── SKILL.md                   ← THE skill — agents read this
├── README.md
├── install.sh                 ← bash installer for all platforms
├── LICENSE                    ← MIT
├── CONTRIBUTING.md
├── .github/
│   └── workflows/
│       └── validate.yml       ← CI: validates structure, frontmatter, JS syntax
├── references/
│   ├── network.md             ← cy.intercept deep-dive
│   ├── assertions.md          ← complete assertion cheatsheet
│   ← selectors.md            ← stability hierarchy + data-testid guide
│   ├── commands.md            ← full cy.* command reference
│   ├── config.md              ← complete cypress.config.js options
│   ├── ci.md                  ← GitHub Actions, GitLab, CircleCI, Jenkins
│   ├── component-testing.md   ← React/Vue/Angular component tests
│   └── patterns.md            ← visual regression, API testing, multi-tab
└── examples/
    ├── auth-flow.cy.js        ← cy.session auth (complete test suite)
    ├── api-intercept.cy.js    ← network stubbing patterns
    ├── page-objects.cy.js     ← full POM implementation
    └── custom-commands.js     ← drop-in command library + TypeScript types

Why flat? When an agent installs this skill (e.g. to ~/.openclaw/skills/cypress-agent-skill/), the repo root is the skill root. The main SKILL.md lives directly at that path with no subdirectories to navigate. Clean, direct, agent-friendly.

Coverage

Topic	Covered
Installation & Configuration	✅
Selector Strategy (data-testid, ARIA, cy.contains)	✅
Assertions (should, expect, chaining)	✅
Network Stubbing (cy.intercept)	✅
Request/Response Assertions	✅
Authentication (cy.session)	✅
Custom Commands + TypeScript types	✅
Page Object Model	✅
Component Testing (React, Vue, Angular)	✅
Forms, File Upload, Tables	✅
Modals, Alerts, iframes	✅
Local Storage / Cookies	✅
Flake Prevention	✅
CI/CD (GitHub Actions, GitLab, CircleCI, Jenkins)	✅
Parallelization (Cypress Cloud)	✅
Environment Variables	✅
Fixtures & Data Management	✅
Accessibility (cypress-axe)	✅
TypeScript Support	✅
Visual Regression (Percy)	✅
API Testing (cy.request)	✅
Multi-tab Handling	✅
Date/Time Manipulation (cy.clock)	✅
Database Seeding via Tasks	✅
Download Testing	✅
Drag and Drop	✅

How Agents Use This Skill

When an AI agent installs this skill:

Agent reads SKILL.md automatically when you ask it to write or fix Cypress tests
For deep dives, the agent reads specific files from references/ on demand
examples/ files serve as copy-paste-ready templates

The skill uses AgentSkills format with OpenClaw-native metadata for gating (only activates when cypress or npx is on PATH).

Platform Install Paths

Agent	Path	Scope
OpenClaw	`~/.openclaw/skills/cypress-agent-skill/`	Global
Claude Code	`./.claude/skills/cypress-agent-skill/`	Project
Claude Code	`~/.claude/skills/cypress-agent-skill/`	Personal
Codex CLI	`./.agents/skills/cypress-agent-skill/`	Project
Cursor	`./.cursor/skills/cypress-agent-skill/`	Project
Global	`~/.agents/skills/cypress-agent-skill/`	All agents

Contributing

See CONTRIBUTING.md. PRs welcome — real-world patterns only, no toy examples.

License

MIT

README.md

🧪 cypress-agent-skill