AI Development: Integrating Atlassian Jira with Claude Code

If you’re thinking about leveraging AI in your development, you should consider using Claude Code with MCP. Claude Code is Anthropic's official command-line interface that puts the power of Claude AI directly into your development environment. It allows you to get intelligent help with coding tasks, debugging, refactoring, and project management — all without leaving your terminal. Because of this, Claude Code is one of the leading developer tools for AI-integrated development right now.

At the heart of Claude Code's extensibility is the Model Context Protocol (MCP). Think of MCP as a universal translator that allows Claude to speak the language of your favorite tools and services. Whether it's your project management software, documentation platforms, or development tools, MCP creates a bridge that lets Claude interact with these systems as naturally as a human would.

There are hundreds of existing MCP servers available; you can view a list of them at https://github.com/modelcontextprotocol/servers.

For this post, we’ll explain how to integrate Atlassian Jira and Confluence as an MCP for Claude Code. This allows Claude to talk with your Atlassian instance and use the data inside it to better understand your requests.

Atlassian + Claude Code

For developers working in teams, Atlassian's suite of tools forms the backbone of project management and knowledge sharing. However, constantly switching between your code editor, terminal, Jira tickets, and Confluence pages can be a productivity killer. As a developer, your ideal world would be one where you never have to leave the editor and can interact with Atlassian from one of the panels in your integrated development environment (IDE).

By integrating Atlassian with Claude Code through MCP, you create a unified workspace where Claude can:

• Pull information from Jira tickets while you're coding
• Search Confluence documentation to answer your questions
• Create and update tickets based on your development work
• Cross-reference code changes with project requirements
• Help maintain documentation alongside your codebase

Setting Up Atlassian Integration with Claude Code

Here's how to connect your Atlassian workspace to Claude Code using the MCP Atlassian connector.

Step 1: Generate Your Atlassian API Token

First, you'll need to create an API token allowing Claude to securely access your Atlassian workspace.

1. Log in to your Atlassian account and navigate to API Token Management.
2. Create a new API Token.
3. Copy the token to a secure temporary location (you'll need it in the next step).

Step 2: Configure the MCP Server

The MCP Atlassian connector runs as a Docker container, making setup straightforward across different operating systems. You'll need to configure it with your Atlassian credentials and workspace URLs.

Create the following JSON configuration, replacing the placeholder values with your actual information:

Unlike many MCP tutorials you might find online, Claude Code doesn't require wrapping this configuration in additional mcpServer or server name objects. The JSON above is all you need to provide. Replace the placeholder values with your Atlassian URL, email address, and API key from the previous step.

Step 3: Add the MCP Server to Claude Code

With your configuration ready, it's time to register the MCP server with Claude Code:

1. Open your terminal.
2. Run the following command:

claude mcp add mcp-atlassian -s user 'PASTE_YOUR_FULL_JSON_HERE'

The -s user flag saves this server configuration across your entire system rather than just the current project, making it available whenever you use Claude Code on your machine.

You should see an "MCP server added" success message confirming the setup.

Step 4: Start Using Atlassian + Claude Code

You are now able to use Claude and Atlassian together from anywhere on your computer, right from your editor:

1. Navigate to any project directory.
2. Start Claude Code by running claude.
3. Type /mcp to see your available MCP servers — you should see mcp-atlassian in the list.
4. Grant Claude access to the tools when prompted (or it will ask for permission each time).

What You Can Do Now

With everything configured, you can now ask Claude questions like:

• "What's the current status of ticket PROJ-123?"
• "Search our Confluence for documentation about the SSO authentication flow for OAuth in Drupal."
• "Create a new Jira ticket for the bug I just found in this code."
• "What are the acceptance criteria for the feature I'm working on?"

Claude will query your Atlassian workspace and provide contextual information right in your terminal, making your development process more informed and efficient. By breaking down the barriers between development tools and project management platforms, you can focus more on solving problems and less on juggling different interfaces.

Putting it All Together

Here is a quick video demonstrating the MCP integration for Claude Code and Jira. I ask it about a ticket and ask it to review the description and comments, and then look at my custom Drupal modules and give me an explanation of the cause and potential solution: https://www.youtube.com/watch?v=z2qvWcwDYnc.

Claude Code was able to successfully diagnose the what, where, and why.

Whether you're debugging a complex issue, implementing a new feature, or trying to understand the context behind a piece of code, having Claude Code connected to your Atlassian workspace means your AI assistant can ride shotgun and help you be more efficient.

Give it a try today and experience the difference of having AI-integrated development. Want more help leveraging Claude Code in your development? Contact us. Our experts would be happy to assist you.