AI Driven Development (AIDD) using an MCP Server
We have developed this plugin to AI tools that supports using your AI LLM model to write software using AI-driven software development practices. This first version is available as a plugin to Anthropic Claude's desktop app. There is also a version for use with the skydeck.ai workspace coming soon. You can get this plugin from https://mcp-get.com/packages/mcp-server-aidd.

MCP Server AIDD Documentation

By Skydeck.ai and East Agile

Overview

MCP Server AIDD (AI-Driven Development) is a comprehensive development environment that combines AI assistance with powerful development tools. This guide will help you understand its capabilities and get started with AI-driven development. Get this from https://mcp-get.com/packages/mcp-server-aidd

Core Capabilities

1. File System Operations

// Read files
const content = await window.fs.readFile('path/to/file', { encoding: 'utf8' });

// Write files
await window.fs.writeFile('path/to/file', content);

// Create directories
await create_directory('new_project');

2. Code Analysis

Using tree-sitter, analyze code structure and relationships:

# Analyze repository structure
analysis = tree_sitter_map(path="./src")

# Results include:
- Function definitions
- Class hierarchies
- Module dependencies
- Code organization patterns

3. Multi-Language Code Execution

Execute code in various languages directly:

# Python
execute_code(
    language="python",
    code="""
    def hello():
        print("Hello from Python!")
    hello()
    """
)

# JavaScript, Ruby, PHP, Go, and Rust also supported

4. Git Integration

Full Git workflow support:

# Initialize repository
git_init(path="project_path")

# Stage and commit
git_add(repo_path="path", files=["file1.py", "file2.py"])
git_commit(repo_path="path", message="Initial commit")

# View changes
git_status(repo_path="path")
git_diff(repo_path="path", target="main")

5. System Integration

# Get system information
system_info = get_system_info()

# Send system notifications
sendNotification(
    title="Task Complete",
    message="Your build has finished"
)

6. Directory Management

# List directory contents
contents = list_directory("./src")

# Search for files
results = search_files(
    path="./src",
    pattern="*.py",
    exclude_patterns=[".git/*", "venv/*"]
)

# Get file information
info = get_file_info("path/to/file.py")

7. Shell Script Execution

execute_shell_script(
    script="""
    echo "Current directory:"
    pwd
    ls -la
    """
)

Advanced Features

1. Browser Automation

# Navigate to URLs
puppeteer_navigate(url="https://example.com")

# Take screenshots
puppeteer_screenshot(
    name="page.png",
    selector=".main-content"
)

# Interact with pages
puppeteer_click(selector="#submit-button")
puppeteer_fill(selector="#search", value="query")

2. External Services

# Maps integration
coordinates = maps_geocode(address="123 Main St")
places = maps_search_places(query="coffee shops")
route = maps_directions(
    origin="Start Point",
    destination="End Point",
    mode="driving"
)

# Brave search
results = brave_web_search(query="python development")
local = brave_local_search(query="restaurants near me")

3. Time Management

# Get current time
time = get_current_time(timezone="America/New_York")

# Convert between timezones
converted = convert_time(
    source_timezone="America/New_York",
    target_timezone="Asia/Tokyo",
    time="14:30"
)

Project Organization

Directory Structure

project/
├── src/           # Source code
├── tests/         # Test files
├── docs/          # Documentation
└── .git/          # Git repository

Best Practices

  1. File Organization
    • Use clear directory structures
    • Keep related files together
    • Follow language-specific conventions
  2. Version Control
    • Regular commits
    • Clear commit messages
    • Feature branches
  3. Code Analysis
    • Regular code structure analysis
    • Monitor dependencies
    • Track code metrics

System Requirements

  • Operating System: Cross-platform support
  • Python: Version 3.12+
  • Node.js: Latest LTS version
  • Git: Latest stable version

Security

  • File operations restricted to allowed directories
  • Secure execution environment
  • Permission-based access control

Example Workflows

1. Setup New Project

# Create project structure
create_directory("new_project")
git_init(path="new_project")

# Add initial files
write_file(
    path="new_project/README.md",
    content="# New Project\n\nDescription here."
)

2. Code Analysis Workflow

# Analyze codebase
analysis = tree_sitter_map(path=".")

# Search for specific patterns
results = search_files(
    path=".",
    pattern="*.py"
)

# Get file details
for file in results:
    info = get_file_info(file)

3. Development Workflow

# Execute tests
execute_code(
    language="python",
    code="pytest tests/"
)

# Commit changes
git_add(repo_path=".", files=["modified_file.py"])
git_commit(repo_path=".", message="Update tests")

Support

Documentation

  • Detailed API reference
  • Example projects
  • Best practices guides

Contact

  • Technical Support: support@skydeck.ai
  • Professional Services: services@eastagile.com

Updates and Maintenance

  • Regular feature updates
  • Security patches
  • Performance improvements

MCP Server AIDD is a product of Skydeck.ai and East Agile, empowering developers with AI-driven development tools and workflows.

© 2025 East Agile. All rights reserved.

Tags
Tips & Tricks
AI
61
Share this
Leave a comment
There are no comments about this article, let us know what you think?

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.