PICT Test Designer - Claude Skill

A Claude skill for designing comprehensive test cases using PICT (Pairwise Independent Combinatorial Testing). This skill enables systematic test case design with minimal test cases while maintaining high coverage through pairwise combinatorial testing.

🎯 What is PICT?

PICT (Pairwise Independent Combinatorial Testing) is a combinatorial testing tool developed by Microsoft. It generates test cases that efficiently cover all pairwise combinations of parameters while drastically reducing the total number of tests compared to exhaustive testing.

Example: Testing a system with 8 parameters and 3-5 values each:

Exhaustive testing: 25,920 test cases
PICT pairwise testing: ~30 test cases (99.88% reduction!)

🚀 Features

Automated Test Case Generation: Converts requirements into structured PICT models
Constraint-Based Testing: Applies business rules to eliminate invalid combinations
Expected Output Generation: Automatically determines expected results for each test case
Comprehensive Coverage: Ensures all pairwise parameter interactions are tested
Multiple Domains: Works for software functions, APIs, web forms, configurations, and more

🔧 Installation

Prerequisites

Claude Code CLI or Claude Code Desktop
(Optional) Python 3.7+ with pypict for advanced usage

Installation Methods

Claude Code skills can be installed via the plugin marketplace or manually by placing them in the .claude/skills/ directory.

Method 1: Install via Claude Code Plugin Marketplace (Easiest) 🌟

Install directly through Claude Code's plugin system:

BASH

# Add the marketplace
/plugin marketplace add omkamal/pypict-claude-skill

# Install the plugin
/plugin install pict-test-designer@pypict-claude-skill

This automatically installs the skill and keeps it updated. The skill will be available across all your projects.

Method 2: Install from GitHub (Manual)

For Personal Use (All Projects):

BASH

# Clone the repository to your personal skills directory
git clone https://github.com/omkamal/pypict-claude-skill.git ~/.claude/skills/pict-test-designer

# Restart Claude Code to load the skill
# The skill will now be available in all your projects

For Project-Specific Use:

BASH

# From your project directory
git clone https://github.com/omkamal/pypict-claude-skill.git .claude/skills/pict-test-designer

# Add to .gitignore if you don't want to commit it
echo ".claude/skills/" >> .gitignore

# Or commit it to share with your team
git add .claude/skills/pict-test-designer
git commit -m "Add PICT test designer skill"

If you want to share this skill with your team via version control:

BASH

# From your project directory
git submodule add https://github.com/omkamal/pypict-claude-skill.git .claude/skills/pict-test-designer
git commit -m "Add PICT test designer skill as submodule"

# Team members clone with:
git clone --recurse-submodules <your-repo-url>

# Or if already cloned:
git submodule update --init --recursive

Method 4: Download Minimal Package from Releases

Download the pre-packaged minimal installation from GitHub Releases:

BASH

# Download the latest minimal package from releases
wget https://github.com/omkamal/pypict-claude-skill/releases/latest/download/pict-test-designer-minimal.zip

# Extract and install for personal use
unzip pict-test-designer-minimal.zip
mv pict-test-designer-minimal ~/.claude/skills/pict-test-designer

# Or for project-specific use
unzip pict-test-designer-minimal.zip
mv pict-test-designer-minimal .claude/skills/pict-test-designer

What's included: SKILL.md, LICENSE, references/ (syntax and examples) What's excluded: Full examples, helper scripts, extended documentation Size: ~9 KB | Latest Version: See Releases

Method 5: Download Full Repository

Download the repository as a ZIP from GitHub
Extract to the skills directory:

BASH

# For personal use (all projects)
unzip pypict-claude-skill-main.zip
mv pypict-claude-skill-main ~/.claude/skills/pict-test-designer

# For project-specific use
unzip pypict-claude-skill-main.zip
mv pypict-claude-skill-main .claude/skills/pict-test-designer

Verify Installation

After installation, restart Claude Code. The skill will load automatically when relevant. You can verify by asking Claude:

Do you have access to the pict-test-designer skill?

Or simply start using it:

Design test cases for a login function with username, password, and remember me checkbox.

🚀 Quick Start

Once installed, you can use the skill in Claude by simply asking:

Design test cases for a login function with username, password, and remember me checkbox.

Claude will:

Analyze the requirements
Identify parameters and values
Generate a PICT model with constraints
Create test cases with expected outputs
Present results in a formatted table

📊 Example: ATM System Testing

This repository includes a complete real-world example of testing an ATM system. See the examples directory for:

ATM Specification: Complete ATM system specification with 11 sections covering hardware, software, security, and functional requirements
ATM Test Plan: Comprehensive test plan generated using PICT methodology with 31 test cases (reduced from 25,920 possible combinations)

ATM Example Summary

System Parameters:

Transaction Types (5): Withdrawal, Deposit, Balance Inquiry, Transfer, PIN Change
Card Types (3): EMV Chip, Magnetic Stripe, Invalid
PIN Status (4): Valid, Invalid attempts 1-3
Account Types (3): Checking, Savings, Both
Transaction Amounts (4): Within limits, at max, exceeds transaction, exceeds daily
Cash Availability (3): Sufficient, Insufficient, Empty
Network Status (3): Primary, Backup, Disconnected
Card Condition (3): Good, Damaged, Expired

Test Results:

Total possible combinations: 25,920
PICT test cases generated: 31
Reduction: 99.88%
Coverage: All pairwise (2-way) interactions
Test execution time: Reduced from weeks to hours

Running the ATM Example

BASH

# In Claude Code
Ask: "Use the pict-test-designer skill to analyze the ATM specification 
in examples/atm-specification.md and generate test cases"

🔍 How It Works

1. Requirements Analysis

Claude analyzes your requirements to identify:

Parameters: Input variables, configuration options, environmental factors
Values: Possible values using equivalence partitioning
Constraints: Business rules and dependencies
Expected Outcomes: What should happen for different combinations

2. PICT Model Generation

Creates a structured model:

# Parameters
Browser: Chrome, Firefox, Safari
OS: Windows, MacOS, Linux
Memory: 4GB, 8GB, 16GB

# Constraints
IF [OS] = "MacOS" THEN [Browser] <> "IE";
IF [Memory] = "4GB" THEN [OS] <> "MacOS";

3. Test Case Generation

Generates minimal test cases covering all pairwise combinations:

Test #	Browser	OS	Memory	Expected Output
1	Chrome	Windows	4GB	Success
2	Firefox	MacOS	8GB	Success
3	Safari	Linux	16GB	Success
...	...	...	...	...

4. Expected Output Determination

For each test case, Claude determines the expected outcome based on:

Business requirements
Code logic
Valid/invalid combinations

🎯 Use Cases

Software Testing

Function testing with multiple parameters
API endpoint testing
Database query testing
Algorithm validation

Configuration Testing

System configuration combinations
Feature flag testing
Environment setup validation
Browser compatibility testing

Web Application Testing

Form validation
User authentication flows
E-commerce checkout processes
Shopping cart functionality

Mobile Testing

Device and OS combinations
Screen size and orientation
Network conditions
App permissions

Hardware Testing

Device compatibility
Interface testing
Protocol validation
Performance under different conditions

📚 Documentation

SKILL.md: Complete skill documentation with workflow and best practices
PICT Syntax Reference: Complete syntax guide (to be created)
Examples: Real-world examples across domains (to be created)
Helper Scripts: Python utilities for PICT (to be created)

💡 Tips for Best Results

Good Parameter Names

✅ Use descriptive names: AuthMethod, UserRole, PaymentType ✅ Apply equivalence partitioning: FileSize: Small, Medium, Large ✅ Include boundary values: Age: 0, 17, 18, 65, 66 ✅ Add negative values: Amount: ~-1, 0, 100, ~999999

Writing Constraints

✅ Document rationale: # Safari only available on MacOS ✅ Start simple, add incrementally ✅ Test constraints work as expected

Expected Outputs

✅ Be specific: "Login succeeds, user redirected to dashboard" ❌ Not vague: "Works" or "Success"

🙏 Credits

This skill is built upon the excellent work of:

Microsoft PICT: The original Pairwise Independent Combinatorial Testing tool developed by Microsoft Research
pypict: Python binding for PICT by Kenichi Maehashi
Community Contributors: All contributors who have helped improve PICT tools

About PICT

PICT was developed by Jacek Czerwonka at Microsoft Research. It's a powerful combinatorial testing tool that has been used extensively within Microsoft for testing complex systems with multiple interacting parameters.

References: