AI Development Pipeline

You run a local and cloud MCP server that connects Claude AI, VSCode, Augment, and cloud services to perform file operations, shell commands, and AI-assisted tasks. This setup lets you work offline with a local server and also deploy a scalable HTTP endpoint for cloud integration, all while exposing a consistent set of MCP tools to automate your development workflow.

How to use

Start the local MCP server to enable direct stdio integration with Claude Desktop. You can also run the cloud MCP server to expose an HTTP endpoint for web-based Claude integration. Use Claude to issue commands that read, write, run shell commands, and run AI-assisted prompts against your project.

How to install

Prerequisites: Node.js 18+ and npm, TypeScript and ts-node, Claude Desktop for local integration, and a Vercel account for cloud deployment.

# 1. Clone the repository
git clone https://github.com/yourusername/ai-development-pipeline-mcp.git
cd ai-development-pipeline-mcp

# 2. Install dependencies
npm install

# 3. Configure environment variables
cp .env.example .env
# Edit .env with your API keys and configuration

Additional setup details

Create a .env file in the root directory with the necessary configuration values for Vercel, Airtable, Square, and analytics. The following variables are used in production and local runs.

# Vercel Configuration
VERCEL_TOKEN=your_vercel_token_here
VERCEL_PROJECT_ID=your_project_id_here

# Airtable Configuration  
AIRTABLE_API_KEY=your_airtable_api_key_here
AIRTABLE_BASE_ID=your_base_id_here
AIRTABLE_TABLE_NAME=your_table_name_here

# Square Configuration
SQUARE_APPLICATION_ID=your_square_app_id_here
SQUARE_ACCESS_TOKEN=your_square_access_token_here

# Analytics Configuration
ANALYTICS_SECRET=your_analytics_secret_here
NEXT_PUBLIC_APP_URL=https://your-app-url.vercel.app

Claude Desktop Integration

Configure Claude Desktop to connect to the MCP server by adding an MCP server entry that points to your local runtime using the standard CLI invocation.

{
  "mcpServers": {
    "ai-development-pipeline": {
      "command": "npx",
      "args": ["ts-node", "/path/to/your/project/local-mcp-server.ts"],
      "env": {}
    }
  }
}

Cloud Deployment (Vercel)

Deploy the cloud MCP server to Vercel to provide an HTTP endpoint for Claude integration.

# Automatic deployment is recommended when connected to GitHub
# 1. Connect to GitHub, create a new project, and import your repository
# 2. Vercel will automatically detect and deploy the project

# Manual deployment (if needed)
npm install -g vercel
vercel deploy --prod

Available tools

read_project_file

Read files from the workspace to retrieve code, configs, or content needed for further processing.

write_project_file

Write or update files within the workspace to apply changes generated by MCP-driven automation.

run_shell_command

Execute shell commands (such as npm, git, or system utilities) from within the MCP context.

check_file_exists

Check whether a specific file path exists in the workspace.

list_directory_files

List the contents of a directory to inspect project structure.

run_augment_prompt

Send prompts to the Augment coding agent for AI-assisted coding guidance.

run_project_tests

Execute project tests to validate changes and ensure stability.