home / mcp / paelladoc mcp server

PAELLADOC MCP Server

AI-first development framework

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "jlcases-paelladoc": {
      "command": "/absolute/path/to/.paelladoc_venv/bin/python",
      "args": [
        "-m",
        "paelladoc.ports.input.mcp_server_adapter",
        "--stdio"
      ],
      "env": {
        "PYTHONPATH": "/path/to/your/project/directory/src:/path/to/your/project/directory",
        "PAELLADOC_DB_PATH": "/path/to/your/project/directory/paelladoc_memory.db"
      }
    }
  }
}

PAELLADOC uses the Model Context Protocol (MCP) to let large language models interact with its AI-first development tools through a standardized, tool-like interface. This enables you to drive documentation projects, manage context and memory, and receive living, context-aware guidance from your LLM client. With MCP, you can start new projects, continue existing ones, verify coverage, and generate documentation interactively, all while preserving intent and history across steps.

How to use

You connect your MCP client (for example, Cursor) to PAELLADOC, then issue high-level commands in natural language. Start a new documentation project with the PAELLADOC commands exposed through MCP, or continue working on an existing project. Your LLM prompts PAELLADOC to create the necessary context, update living documentation, and guide you through template selection, memory configuration, and artifact generation. The workflow emphasizes intent preservation, context, and collaborative human–AI decision making.

How to install

Prerequisites: ensure you have Python 3.12 or later installed on your system.

1. Create and activate a dedicated Python virtual environment to isolate PAELLADOC dependencies.

python3.12 -m venv .paelladoc_venv

# Activate the environment
# On macOS/Linux (bash or zsh):
source .paelladoc_venv/bin/activate

# On Windows (PowerShell):
.\.paelladoc_venv\Scripts\activate.ps1

2. Install PAELLADOC into the activated virtual environment.

pip install paelladoc

3. Configure the database path for PAELLADOC memory storage. You can set it via environment or via your MCP tool configuration.

# Option 1: Environment variable (less reliable for MCP integration)
export PAELLADOC_DB_PATH="$HOME/.paelladoc/memory.db"

# Option 2: MCP configuration (recommended for MCP integration)
# You will place this in your MCP JSON and point to a DB path in your project.

4. Configure your MCP client to launch PAELLADOC in stdio mode using the dedicated Python interpreter inside your virtual environment. Use the example below as a guide to create your MCP server entry.

Configuration example for MCP client

{
  "mcpServers": {
    "Paelladoc": {
      "command": "/absolute/path/to/.paelladoc_venv/bin/python", 
      "args": [
        "-m",
        "paelladoc.ports.input.mcp_server_adapter",
        "--stdio"
      ],
      "cwd": "/path/to/your/project/directory",
      "env": {
        "PAELLADOC_DB_PATH": "/path/to/your/project/directory/paelladoc_memory.db",
        "PYTHONPATH": "/path/to/your/project/directory/src:/path/to/your/project/directory"
      },
      "disabled": false
    }
  },
  "mcp.timeout": 120000
}

Notes on usage and workflow

Once connected, your LLM can call the following MCP commands to manage PAELLADOC work: start new projects, list existing ones, select a project to load its memory, continue work on a project, request help, view a basic status, and verify documentation. The system is designed to preserve intent and context across steps, so each action records the rationale and context behind decisions.

Security and troubleshooting

Keep your MCP access isolated and use secure environments for memory storage. If you encounter environment-variable inheritance issues when the LLM launches PAELLADOC, prefer configuring PAELLADOC_DB_PATH directly in the MCP client configuration and ensure the Python virtual environment is reachable by the MCP process.

Examples and quick start

In your MCP-enabled chat interface, you can simply say or prompt: PAELLA to begin a new documentation project. The system will guide you through project details, template choices, and how to structure your documentation in a living, context-preserving manner.

Available tools

ping

Health check to verify the server is reachable and responsive.

paella_init

Initialize a new PAELLADOC project with base path, language, and project name.

paella_list

List existing PAELLADOC projects stored in memory.

paella_select

Select a project to load its memory and continue work.

core_continue

Continue work on a selected project and load its memory with suggested next steps.

core_help

Provide help information about available MCP commands and usage.

core_list_projects

List existing PAELLADOC projects (alternative to paella_list).

core_verification

Check documentation quality and completeness for the current project.