Plano

Plano is a minimalist, powerful project management system designed with one core philosophy: calm, structured oversight over everything. It features built-in visual diagrams alongside standard task tracking so your architecture, process flows, and daily tasks all live in one cohesive environment.

Features

• Project & Task Management: A full project tracking system with milestones, tasks (kanban/list views), tags, and updates.
• Dynamic Visual Boards: Create interactive diagrams directly within projects (Process Flows, DB Diagrams, Flowcharts, Idea Maps, and more).
• Relational DB Visualization: Rich metadata schema for database nodes, including columns, types, and primary/foreign keys.
• MCP Native Integration: Built from the ground up to be easily operated via the Model Context Protocol (MCP).
• Real-time Reactivity: Live system-wide updates and AI tool-call indicators powered by Server-Sent Events (SSE).
• Global Portability: High-fidelity JSON export/import for entire projects including diagrams, tasks, and milestones.
• Automatic File Tracking: Automatically captures and persists exactly which files were modified in every task using git diff.
• API Tester: Full built-in API testing interface.

Screenshots

Quick Start

1. Setup Environment

Linux / macOS:

python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Windows:

python -m venv venv
.\venv\Scripts\activate
pip install -r requirements.txt

2. Start the Server

Linux / macOS:

./start.sh

Windows:

python run.py

3. Usage

• Dashboard: http://localhost:5000/dashboard
• Health Check: http://localhost:5000/health
• API Docs: http://localhost:5000/api

4. Recommended: Install AI Agent Directives

For the best experience, we highly recommend installing the pre-configured Directives system. This gives AI agents (like Cursor, Windsurf, or Antigravity) structured oversight of your project.

To install:

1. Copy the contents of AGENTS.md.
2. Paste them into your agent's custom instructions file (e.g., .cursorrules, .windsurfrules, or project settings).
3. The agent will now automatically use Plano to track its own progress seamlessly.

MCP Installation

Plano includes a native MCP (Model Context Protocol) server that allows AI assistants to directly manage your projects and diagrams.

Quick Install (AI-Driven)

If you are using an AI assistant that supports MCP auto-discovery (like Antigravity or Windsurf), copy the absolute path to this project and tell the AI:

{{PATH}} install this mcp tool

Manual Configuration (Claude Desktop)

Add the following to your claude_desktop_config.json (replacing /PATH/TO/PLANO with the absolute path to this project):

Linux / macOS:

{
  "mcpServers": {
    "plano": {
      "command": "/PATH/TO/PLANO/venv/bin/python",
      "args": ["/PATH/TO/PLANO/mcp_server.py"]
    }
  }
}

Windows:

{
  "mcpServers": {
    "plano": {
      "command": "C:\\PATH\\TO\\PLANO\\venv\\Scripts\\python.exe",
      "args": ["C:\\PATH\\TO\\PLANO\\mcp_server.py"]
    }
  }
}

Note: The MCP server automatically manages the lifecycle of the Flask backend. On the first tool call, it will start the backend on port 5050 if it isn't already running.

Recent Changes (Changelog)

• v2.2.0 (2026-03-21):

  • Lean Core: Removed GitHub integration in favor of a simpler, more robust local-first architecture using a standalone git diff capture tool.
  • DB Schema Visualization: Implemented rich metadata for database nodes, enabling display and management of columns, types, and primary/foreign keys within diagrams.
  • UI/UX Refinement: Widened Kanban board columns for better readability and added "AI Working" status badges to project dashboards.
  • Timezone Support: Added full support for local timezone rendering across all UI timestamps.
  • Layout Engine: Improved auto-layout spacing and added default Left-to-Right orientation for process flows.

• v2.1.0 (2026-03-20):

  • Real-time Engine: Implemented global SSE for live UI updates and AI activity indicators.
  • QA Workflow: Added dedicated Bugs triage and Testing columns to the task board.
  • Data Portability: Launched the high-fidelity Export/Import engine for project backups and migrations.
  • Visual Board Fixes: Improved diagram layout performance and node visibility.

Architecture

Plano is built using a lightweight stack:

• Backend: Flask / Python
• Database: SQLite & SQLAlchemy (plano.db)
• Frontend: Vanilla JS / CSS / HTML with no heavy build systems or JS frameworks.
• MCP: Built on FastMCP for quick conversational AI usage.

License

Copyright (C) 2026 nopan-studio MIT License See LICENSE for details.