PictMCP

takeyaqa/PictMCP

0 starsv0.1.0STDIORegistry activeMITUpdated 2026-03-22Community

Deploy on Hostinger →Free API →AI & MLaiView on GitHubGitHub repository npm package Report an issue

Is this your server?

Add your score badge to your README and get your server in front of 45k+ builders a month.

Get your badge →Submit updates →Become a Verified Publisher →

Works with

Claude CodeClaude DesktopCursorVS CodeClineCodex CLIOpenClaw+ any MCP client

Install to Claude Code

claude mcp add pictmcp -- npx -y pictmcp

Summary

Provides pairwise test generation for AI assistants using the PICT algorithm, running locally via WebAssembly.

Connect from your MCP client

One-click install

Add this server to your editor with a single click. Fill in any required credentials afterward.

Install in Cursor Install in VS Code

Claude Code

Run this once and Claude Code registers the server for you:

claude mcp add pictmcp -- npx -y pictmcp

Claude Desktop

Add this to claude_desktop_config.json under Settings → Developer → Edit Config:

{
  "mcpServers": {
    "pictmcp": {
      "command": "npx",
      "args": [
        "-y",
        "pictmcp"
      ]
    }
  }
}

Cursor

Add this to .cursor/mcp.json in your project (or ~/.cursor/mcp.json for all projects):

{
  "mcpServers": {
    "pictmcp": {
      "command": "npx",
      "args": [
        "-y",
        "pictmcp"
      ]
    }
  }
}

Cline and other MCP clients

Most MCP clients accept the standard mcpServers JSON block:

{
  "mcpServers": {
    "pictmcp": {
      "command": "npx",
      "args": [
        "-y",
        "pictmcp"
      ]
    }
  }
}

Codex CLI

codex mcp add pictmcp -- npx -y pictmcp

# or add to ~/.codex/config.toml:
[mcp_servers.pictmcp]
command = "npx"
args = ["-y", "pictmcp"]

OpenClaw

OpenClaw reads MCP servers from the mcp.servers section of ~/.openclaw/openclaw.json (managed via `openclaw mcp add` or the mcporter skill):

{
  "mcp": {
    "servers": {
      "pictmcp": {
        "command": "npx",
        "args": [
          "-y",
          "pictmcp"
        ]
      }
    }
  }
}

README.md

PictMCP

[!CAUTION] This package has been archived and will no longer be maintained. Please consider using takeyaqa/tester-skills.

Pairwise testing for your AI assistant

PictMCP is an MCP server for software developers who design test cases with AI assistants, providing reliable, algorithm-correct pairwise test generation.

Why use this?

AI is great at test design, but not at combinatorial math.
Pairwise generation must be deterministic and correct.
PictMCP separates thinking (AI) from calculation (PICT).

Prefer a GUI? Check out PictRider.

Features

🔒 Local Processing - All processing runs locally with no external network calls
⚡ WebAssembly Powered - Fast execution using Microsoft's PICT algorithm compiled to WebAssembly
🔗 Constraint Support - Define constraints to filter out invalid parameter combinations
📊 Structured Output - Returns well-structured JSON results for easy integration

Installation

Prerequisites

Node.js (v22 or higher)

MCP Client Configuration

Add the following configuration to your MCP client. This is an example configuration; the exact format may vary depending on your client. Please refer to your MCP client's documentation for details.

{
  "mcpServers": {
    "PictMCP": {
      "command": "npx",
      "args": ["-y", "pictmcp"]
    }
  }
}

Quick Start

Once installed, you can ask your AI assistant to generate test cases using pairwise combinatorial testing.

Example Prompt

Generate test cases for a login form with the following parameters: - Browser: Chrome, Firefox, Safari - OS: Windows, macOS, Linux - Language: English, Japanese, Spanish

The AI assistant will use the generate-test-cases tool to create an optimized set of test cases that covers all pairwise combinations.

Example Result

AI assistants typically format the results as a table:

| # | Browser | OS | Language | | --- | ------- | ------- | -------- | | 1 | Chrome | Linux | Japanese | | 2 | Chrome | macOS | Spanish | | 3 | Safari | Linux | Spanish | | 4 | Firefox | Linux | English | | 5 | Safari | Windows | English | | 6 | Firefox | Windows | Spanish | | 7 | Firefox | macOS | Japanese | | 8 | Safari | macOS | Japanese | | 9 | Chrome | macOS | English | | 10 | Chrome | Windows | Japanese |

Example with Constraints

Generate test cases for: - Browser: Chrome, Firefox, Safari - OS: Windows, macOS, Linux - Language: English, Japanese, Spanish With constraint: Safari only works on macOS

You can describe constraints in plain language — the AI assistant will convert them into PICT constraint syntax automatically.

| # | Browser | OS | Language | | --- | ------- | ------- | -------- | | 1 | Firefox | Linux | Spanish | | 2 | Chrome | Windows | Spanish | | 3 | Firefox | Windows | Japanese | | 4 | Chrome | Linux | Japanese | | 5 | Chrome | macOS | English | | 6 | Firefox | Windows | English | | 7 | Chrome | Linux | English | | 8 | Safari | macOS | Spanish | | 9 | Safari | macOS | Japanese | | 10 | Firefox | macOS | Spanish | | 11 | Safari | macOS | English |

FAQ

Does this communicate with external servers?

No. All processing runs locally with no external network calls.

I already use the `pict` CLI. Do I need this?

If your AI agent can execute CLI commands directly, you may not need this tool. However, PictMCP provides:

A standardized MCP interface for AI assistants
No need to install PICT separately (WebAssembly-based)
Structured JSON output instead of TSV

What is pairwise testing?

Pairwise testing (also known as all-pairs testing) is a combinatorial testing method that generates test cases covering all possible pairs of input parameters. This significantly reduces the number of test cases while maintaining high defect detection rates.

What constraint syntax is supported?

You don't need to write PICT syntax directly. Simply describe constraints in natural language and your AI assistant will handle the conversion. PictMCP supports the full PICT constraint syntax. See the PICT documentation for details.

License

This project is licensed under the MIT License—see the LICENSE file for details.

Disclaimer

PictMCP is provided "as is", without warranty of any kind. The authors are not liable for any damages arising from its use.

Generated test cases do _not guarantee complete coverage_ or _the absence of defects_. Please supplement pairwise testing with other strategies as appropriate.

PictMCP is an independent project and is not affiliated with Microsoft Corporation.

---

If you find PictMCP useful, please consider starring the repository.

See related servers & alternatives →

Score

0–100

57/ 100

Grade

Popularity2/25

No GitHub stars yet.

Maintenance14/20

Updated within the last 6 months.

Completeness20/20

Complete metadata: install, license, description, config, and tags.

Trust9/15

Listed and active in the MCP registry.

Security12/20

Tool surface not yet analyzed.

Scored automatically from popularity, maintenance, completeness, trust, and security. How we score →

Proud of your score? Add this badge to your README.

Paste a snippet into your GitHub README. The badge updates automatically and links back to this page.

Score badge

Markdown

[![PictMCP MCP server](https://www.claudemarket.ai/mcp/takeyaqa/PictMCP/badges/score.svg)](https://www.claudemarket.ai/mcp/takeyaqa/PictMCP)

HTML

<a href="https://www.claudemarket.ai/mcp/takeyaqa/PictMCP"><img src="https://www.claudemarket.ai/mcp/takeyaqa/PictMCP/badges/score.svg" alt="PictMCP MCP server"/></a>

Card badge

Markdown

[![PictMCP MCP server](https://www.claudemarket.ai/mcp/takeyaqa/PictMCP/badges/card.svg)](https://www.claudemarket.ai/mcp/takeyaqa/PictMCP)

HTML

<a href="https://www.claudemarket.ai/mcp/takeyaqa/PictMCP"><img src="https://www.claudemarket.ai/mcp/takeyaqa/PictMCP/badges/card.svg" alt="PictMCP MCP server"/></a>