A MCP implementation of the personal intelligence framework (PIF)
This project implements the Model Context Protocol (MCP) as a practical embodiment of the Personal Intelligence Framework (PIF). Through structured tools and progressive interaction patterns, it creates spaces for meaningful development of understanding between humans and AI.
Note: This implementation has been tested on both Windows and macOS/Linux systems.
git clone [https://github.com/hungryrobot1/MCP-PIF]
cd mcp-pif
npm install
Configure the server:
MCP_WORKSPACE_ROOT environment variable to specify a workspace locationMCP_CONFIG environment variable with a JSON string of configuration optionssrc/config.ts to modify the default configurationBuild the server:
npm run build
Configure Claude Desktop Client:
claude_desktop_config.json:{
"mcpServers": {
"mcp-pif": {
"command": "node",
"args": ["path/to/your/mcp-pif/build/index.js"],
"cwd": "path/to/your/mcp-pif",
"env": {}
}
}
}
path/to/your/mcp-pif with your actual repository pathConnect Claude Desktop Client:
The server will create and manage the following structure in your configured workspace:
workspace/
├── home/
│ ├── meta/
│ │ └── journal/ # For storing journal entries
│ └── projects/ # For user projects
The implementation provides a set of core tools designed to support structured interaction:
pwd, cd, read, write, mkdir, delete, move, renamereason: Develop connected insightsthink: Create temporal spaces for contemplationjournal_create: Document developmentsjournal_read: Explore patterns// Create a structured thought pattern
reason: {
thoughts: [
{ content: "Initial observation" },
{
content: "Building on previous thought",
relationType: "sequence",
relationTo: 0
}
]
}
// Document development
journal_create: {
title: "Implementation Pattern",
content: "Insights about development...",
tags: ["development", "patterns"]
}
The MCP-PIF server is designed to work seamlessly on Windows, macOS, and Linux environments:
The system is built around modular tools that create conditions for structured emergence:
src/
├── core/ # Framework foundations
├── mcp_modules/ # Tool implementations
└── api/ # External integrations
Each tool follows consistent patterns while maintaining its unique role:
The PIF represents a new approach to human-AI collaboration based on:
Rather than prescribing fixed patterns, the implementation creates bounded spaces where understanding can emerge through:
Understanding develops through:
The system supports different levels of engagement:
For those primarily interested in practical implementation:
For those interested in extending the system:
For those interested in deeper patterns:
This project welcomes contributions that engage with both implementation and theoretical aspects:
Comprehensive documentation is available:
The project continues to evolve through:
This implementation embodies a view where:
The system is more than a set of tools - it is a space for exploring how human and AI intelligence can develop through structured interaction. Each session is an opportunity to discover new patterns of understanding and collaboration.
