🤖 Bitbucket MCP Server Tutorial

🚀 Build AI-powered code review bots that integrate with your Bitbucket workflow!

A comprehensive tutorial for building a Model Context Protocol (MCP) server that connects AI assistants like Claude Desktop and Cursor to Bitbucket repositories for intelligent code review and repository management.

⭐ Why This Tutorial?

• 🎯 Production-Ready: Complete server with 11 tools and 4 resources
• 📚 Beginner-Friendly: Step-by-step guide with copy-paste code snippets
• 🤖 AI Integration: Works with Claude Desktop, Cursor, and any MCP-compatible AI
• 🔧 Real-World Usage: Actual PR review automation, not just API demos
• ⚡ Fast Setup: Get running in under 10 minutes

🎯 What You'll Learn

• MCP Fundamentals: Understand the Model Context Protocol and how it connects AI assistants to external tools
• Server Development: Build a production-ready MCP server using FastMCP framework
• API Integration: Connect to Bitbucket's REST API for repository operations
• AI Assistant Integration: Configure Claude Desktop and Cursor to use your MCP server

🚀 What This Server Does

Transform your AI assistant into a powerful development companion that can:

🔧 Repository Management

• 📋 List and explore Bitbucket repositories with intelligent filtering
• 📊 Get detailed repository analytics and metadata
• 🔍 Access repository data through MCP resources for complex queries

🔀 Pull Request Automation

• 📝 List, review, and analyze pull requests automatically
• 💻 Get detailed PR information and complete code diffs
• ⚡ Manage PR workflows (approve, merge, decline) with AI reasoning
• 💬 Add intelligent comments and participate in collaborative reviews

🤖 AI-Powered Code Review

• 🔍 Analyze code changes with context-aware suggestions
• 📈 Identify potential issues, optimizations, and best practices
• 🎯 Generate meaningful code review comments automatically
• 🔄 Streamline your entire review process with AI assistance

Real Example: "Hey Claude, review the latest PR in my-repo and suggest improvements" → Your AI assistant fetches the PR, analyzes the diff, and provides detailed code review feedback!

📋 Prerequisites

• Python 3.8+ (Python 3.9+ recommended)
• Bitbucket Account with API access
• Basic Python Knowledge (variables, functions, async/await)
• Code Editor (VS Code, Cursor, or similar)

🏗️ Project Structure (Tutorial-Ready)

bitbucket-mcp-tutorial/
├── README.md                    # This comprehensive guide
├── LICENSE                     # MIT license
├── mcp_server.py               # Main MCP server (simplified & commented)
├── bitbucket_client.py         # Bitbucket API client
├── test_mcp_server.py          # Test script to verify functionality
├── config_helper.py            # Helper for generating configurations
├── requirements.txt            # Python dependencies
├── .env.example               # Environment variables template
└── docs/
    └── ARCHITECTURE.md         # System design and data flow

⚡ Quick Start (5 Minutes)

1. Clone and Setup

git clone https://github.com/shibyan-ai-engineer/bitbucket-mcp-tutorial
cd bitbucket-mcp-tutorial
pip install -r requirements.txt

2. Configure Environment

cp .env.example .env
# Edit .env with your Bitbucket credentials

3. Test the Server

python test_mcp_server.py --quick

4. Configure AI Assistants

python config_helper.py

🔧 Detailed Setup Guide

Step 1: Python Environment Setup

Option A: Using pip (Recommended for beginners)

# Create project directory
mkdir bitbucket-mcp-tutorial
cd bitbucket-mcp-tutorial

# Install dependencies
pip install -r requirements.txt

Option B: Using virtual environment (Recommended for production)

# Create virtual environment
python -m venv venv

# Activate virtual environment
# On macOS/Linux:
source venv/bin/activate
# On Windows:
venv\\Scripts\\activate

# Install dependencies
pip install -r requirements.txt

Step 2: Bitbucket API Configuration

1. Create App Password:

   • Go to Bitbucket → Settings → Personal settings → App passwords
   • Create new app password with: Repositories (Read, Write), Pull requests (Read, Write)
   • Save the generated password securely

2. Configure Environment Variables:

   cp .env.example .env
   

   Edit .env file:

   BITBUCKET_WORKSPACE=your-workspace-name
   BITBUCKET_USERNAME=your-username
   BITBUCKET_APP_PASSWORD=your-app-password
   

Step 3: Test Your Setup

Quick Test (30 seconds):

python test_mcp_server.py --quick

Full Test (2 minutes):

python test_mcp_server.py

Expected output:

✅ Successfully imported Bitbucket MCP server
✅ Connected successfully!
🔧 Available Tools (11): [list of all tools]
📂 Available Resources (4): [list of all resources]
✅ All tests completed successfully!

Step 4: Configure AI Assistants

For Claude Desktop:

python config_helper.py --claude

For Cursor:

python config_helper.py --cursor

Manual Configuration: The config helper will show you exactly what to add to your AI assistant's configuration files.

🎓 Understanding the Code

Core Components

1. MCP Server (mcp_server.py)

• FastMCP framework setup
• 11 tools for Bitbucket operations
• 4 resources for data access
• Error handling and logging

2. Bitbucket Client (bitbucket_client.py)

• HTTP client for Bitbucket API
• Authentication handling
• Request/response processing

3. Test Script (test_mcp_server.py)

• Comprehensive functionality testing
• Performance benchmarking
• Integration verification

Key Tools Explained

# Tool 1: List Repositories
@mcp.tool
async def list_repositories(role: str = "member"):
    """List repositories by user role"""
    # Implementation details...

# Tool 2: Get Repository Info  
@mcp.tool
async def get_repository_info(repo_slug: str):
    """Get detailed repository information"""
    # Implementation details...

# Tool 3: List Pull Requests
@mcp.tool
async def list_pull_requests(repo_slug: str, state: str = "OPEN"):
    """List pull requests with filtering"""
    # Implementation details...

Resources Explained

# Resource 1: Repositories List
@mcp.resource("bitbucket://repositories")
async def get_repositories_resource():
    """Provide access to repositories data"""
    # Implementation details...

# Resource 2: Specific Repository
@mcp.resource("bitbucket://repo/{repo_slug}")
async def get_repository_resource(repo_slug: str):
    """Provide access to specific repository data"""
    # Implementation details...

🔗 Integration with AI Assistants

Claude Desktop Integration

After running python config_helper.py --claude, add the generated configuration to:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\\Claude\\claude_desktop_config.json

Example configuration:

{
  "mcpServers": {
    "bitbucket": {
      "command": "python",
      "args": ["/absolute/path/to/mcp_server.py"],
      "env": {
        "BITBUCKET_WORKSPACE": "your-workspace",
        "BITBUCKET_USERNAME": "your-username", 
        "BITBUCKET_APP_PASSWORD": "your-app-password"
      }
    }
  }
}

Cursor Integration

After running python config_helper.py --cursor, add the generated configuration to Cursor settings.

📊 Architectural Overview

┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐
│   AI Assistant  │    │   MCP Server     │    │   Bitbucket     │
│                 │    │                  │    │                 │
│  Claude Desktop │◄──►│  11 Tools        │◄──►│  REST API       │
│     Cursor      │    │  4 Resources     │    │  Repositories   │
│                 │    │  FastMCP         │    │  Pull Requests  │
└─────────────────┘    └──────────────────┘    └─────────────────┘

Data Flow

1. User Request: "Review the latest PR in my repository"
2. AI Assistant: Parses request and calls MCP tools
3. MCP Server: Processes tool calls using Bitbucket API
4. Bitbucket API: Returns repository and PR data
5. MCP Server: Formats response for AI assistant
6. AI Assistant: Presents intelligent analysis to user

🛠️ Available Tools & Resources

🔧 Tools (11 total)

Tool	Purpose	Parameters
list_repositories	List user repositories	role (admin/member/contributor)
get_repository_info	Get repo details	repo_slug
list_pull_requests	List PRs	repo_slug, state
get_pull_request_info	Get PR details	repo_slug, pr_id
get_pull_request_diff	Get PR code diff	repo_slug, pr_id
add_pr_comment	Add PR comment	repo_slug, pr_id, content
approve_pr	Approve PR	repo_slug, pr_id
unapprove_pr	Remove approval	repo_slug, pr_id
merge_pr	Merge PR	repo_slug, pr_id, merge_strategy
decline_pr	Decline PR	repo_slug, pr_id, reason
get_pr_comments	Get PR comments	repo_slug, pr_id

📂 Resources (4 total)

Resource	URI Pattern	Purpose
Repositories	bitbucket://repositories	List all repositories
Repository	bitbucket://repo/{repo_slug}	Specific repository data
Pull Requests	bitbucket://repo/{repo_slug}/pullrequests	Repository's PRs
PR Comments	bitbucket://pr/{repo_slug}/{pr_id}/comments	PR comments

🎪 Live Demo Usage Examples

🔥 AI-Powered Code Review in Action

👤 You: "Review the latest PR in my-webapp-project"

🤖 AI Assistant: 
✅ Found PR #42: "Add user authentication system"
📊 Analyzing 15 changed files, 342 additions, 89 deletions...

🔍 Code Review Summary:
• Strong implementation of JWT authentication
• Potential security issue: password validation needs strengthening
• Suggest adding rate limiting to login endpoint
• Missing unit tests for auth middleware
• Database migration looks good

💬 Posted detailed review comment with specific line suggestions!

👤 You: "What repositories need urgent attention?"

🤖 AI Assistant:
📋 Analyzed 12 repositories across your workspace:

🚨 High Priority:
• "mobile-app" - 3 open PRs over 2 weeks old
• "api-service" - Security vulnerability in dependencies

⚠️ Medium Priority:  
• "frontend-dashboard" - 1 large PR awaiting review
• "data-pipeline" - No recent activity, stale issues

✅ All Good:
• "docs-site", "config-service", "monitoring-tools"

🎯 Repository Exploration

👤 You: "What repositories do I have access to in the mobile team workspace?"

🤖 AI Assistant: Found 8 repositories with 'mobile' relevance:
📱 "ios-app" (Swift) - 2.3MB, updated 2 days ago
🤖 "android-app" (Kotlin) - 5.1MB, updated yesterday  
🔧 "mobile-api" (Python) - 1.8MB, updated 3 hours ago
...

🐛 Troubleshooting

Common Issues

1. Import Errors

# Error: ModuleNotFoundError: No module named 'fastmcp'
# Solution: Install dependencies
pip install -r requirements.txt

2. Authentication Errors

# Error: Unauthorized (401)
# Solution: Check .env file configuration
python config_helper.py --test-auth

3. Server Connection Issues

# Error: Connection refused
# Solution: Test server locally first
python test_mcp_server.py --quick

Debug Mode

Run with debug logging:

FASTMCP_DEBUG=1 python mcp_server.py

Run tests with verbose output:

python test_mcp_server.py --verbose

📚 Learning Resources

Next Steps

1. Explore the Code: Read through mcp_server.py with educational comments
2. Try Live Examples: Use your configured AI assistant to interact with repositories
3. Extend Functionality: Add new tools for issues, branches, or commits
4. Build Your Own: Create MCP servers for other APIs (GitHub, GitLab, etc.)

Additional Documentation

• docs/ARCHITECTURE.md - Detailed system design and technical overview

External Resources

• Model Context Protocol Specification
• FastMCP Documentation
• Bitbucket REST API Reference

🤝 Contributing

This tutorial project welcomes improvements! Star ⭐ this repo if it helped you build amazing AI-powered development tools!

🎯 Areas for Contribution:

• 🔧 Additional Bitbucket API integrations (Issues, Deployments, Pipelines)
• 🛡️ Enhanced error handling and retry mechanisms
• 🧪 More comprehensive test coverage
• 📖 Documentation improvements and translations
• 💡 Example use cases and AI prompting strategies
• 🔗 Integration guides for other AI assistants

Join our community of AI-powered developers! 🚀

📄 License

MIT License - Feel free to use this tutorial for learning, teaching, and building awesome AI tools!

---

⭐ Love this project? Give it a star!

🎯 Ready to revolutionize your code review process? Run python test_mcp_server.py --quick to get started!