ten1010-io

Claude Toolkit

By Ten — Claude Code plugins for developer productivity.

Installation

# Step 1: Add to marketplace
/plugin marketplace add ten1010-io/claude-toolkit

# Step 2: Install the plugin you need
/plugin install aqa@ten1010-io          # AI Quality Assurance
/plugin install ship@ten1010-io         # Code shipping & delivery workflow

Plugins

aqa — AI Quality Assurance

AI-driven QA automation that executes YAML test scenarios via browser-use CLI.

/aqa-run

Features:

• YAML-based test scenario definition with cases structure (multiple test cases per file)
• AI-powered element detection (no CSS selectors needed)
• expected_result: "fail" support for error/negative test cases
• Automatic SSL certificate warning bypass
• Parallel execution with worker pool pattern
• Optional screenshot capture (before/after per step)
• HTML report + summary.json generation
• Sensitive data masking in reports
• browser-use CLI auto-detection (global, project venv, home venv)

Usage:

/aqa-run scenarios/auth/login.yaml
/aqa-run scenarios/auth/                       # run all scenarios in directory
/aqa-run scenarios/auth/login.yaml --headless  # headless mode

Options:

Option	Default	Description
--headed	Yes	Run with a visible browser window
--headless	No	Run in headless mode (no browser UI)
--screenshot	Off	Capture before/after screenshots for every step
--parallel N	2	Run N cases concurrently in separate browser sessions

# Fast run (default): no screenshots, 2 cases in parallel
/aqa-run login.yaml

# With screenshots
/aqa-run login.yaml --screenshot

# Run 4 cases in parallel, headless
/aqa-run login.yaml --parallel 4 --headless

# Sequential execution (1 at a time)
/aqa-run login.yaml --parallel 1

Note: When running in parallel, resource-creating values (e.g., project names) are automatically suffixed with _1, _2, etc. to avoid conflicts between concurrent cases.

Scenario format:

name: "Login"
description: "Verify login functionality"
tags: [auth, smoke]

cases:
  - name: "Successful login"
    priority: critical
    expected_result: "pass"
    test_data:
      BASE_URL: "https://example.com"
      username: "testuser"
      password: "secret"
    steps:
      - action: "${BASE_URL}/login 페이지로 이동"
      - action: "아이디 입력란에 ${username} 입력"
      - action: "비밀번호 입력란에 ${password} 입력"
        sensitive: true
      - action: "로그인 버튼 클릭"
      - action: "Dashboard 텍스트가 보이는지 확인"
    cleanup:
      - type: clear_cookies

  - name: "Wrong password"
    priority: high
    expected_result: "fail"
    test_data:
      BASE_URL: "https://example.com"
      username: "testuser"
      [REDACTED:Password]
    steps:
      - action: "${BASE_URL}/login 페이지로 이동"
      - action: "아이디 입력란에 ${username} 입력"
      - action: "비밀번호 입력란에 ${password} 입력"
        sensitive: true
      - action: "로그인 버튼 클릭"
      - action: "에러 메시지가 표시되는지 확인"
    cleanup:
      - type: clear_cookies

Each step only needs the action field — Claude reads the natural language description and determines the appropriate browser commands automatically. All variables including BASE_URL must be defined in test_data.

Legacy single-scenario format (without cases) is also supported for backward compatibility.

/aqa-gen

Interactive scenario generator that creates YAML test files through a guided Q&A process.

Features:

• Step-by-step guided input collection
• BASE_URL automatically saved into test_data (no re-entry needed on re-run)
• Login precondition support (auto-prepends login steps to every case)
• Automatic error case generation based on feature type (login, signup, search, form submission)
• Auto-generated tags from feature name/description
• Multi-language support (responds in the user's language)

Usage:

/aqa-gen

The command will ask you:

1. Feature name (e.g., Login, Signup)
2. Description
3. Login required? (auto-prepends login steps if yes)
4. Target page URL (BASE_URL is extracted and saved automatically)
5. Test data for the success case
6. Steps for the success case
7. Whether to auto-generate error cases
8. Save path

Prerequisites:

• browser-use CLI installed (uv venv + Python 3.12 recommended)

---

ship — PR Workflow Automation

Automates pull request creation with AI-generated descriptions.

/pr

Features:

• Analyzes full commit history (not just latest commit)
• Generates structured PR title and description
• Pushes branch and creates PR via gh CLI
• Conventional commit style PR titles
• Draft PR support

Usage:

/pr                        # PR to main
/pr --base develop         # PR to develop
/pr --draft                # Draft PR to main

Options: