How to connect Jira MCP and Claude Code for effortless project management



This content originally appeared on DEV Community and was authored by Rohith Singh

As a developer, I just get mad when I have to manage my projects manually. When I first started juggling multiple projects – things were okay at first, but as the workload grew, so did the time I spent updating tickets, searching for status reports, and clarifying project requirements. I often found myself wishing for a smarter way to bridge the gap between project management and the development process, where I could just ask an AI agent to do the boring stuff for me.

And if you’re working in a team, you might be facing the same problem.

So… what if we could just ask some AI Agent to handle project management while we focus on the things that actually matter?

Easy Life

In this article, I’ll show you how to connect Jira MCP and Claude Code for effortless project management, right from your terminal, with zero friction.

But First… What is MCP? And What’s Jira MCP?

Think of MCP as a bridge that connects your project management tools to your AI agent. It acts like an adapter, enabling your AI agent (Client) to understand and interact with your tools.

According to Anthropic (the team behind Claude, and MCP),

“MCP is an open protocol that standardizes how applications provide context to LLMs. Think of MCP like a USB-C port for AI applications. Just as USB-C provides a standardized way to connect your devices to various peripherals and accessories, MCP provides a standardized way to connect AI models to different data sources and tools.”

credits: modelcontextprotocol.io

Now, Jira MCP is a Composio-powered MCP server integration for Jira. It lets your AI agent (the client) connect to Jira and interact with your project workflows, including bug tracking, issue tracking, sprint planning, and more. The prompt you pass to your AI agent is converted into a JSON object, which is then sent to the Jira MCP server. The server then interacts with the Jira API to perform the action using the tools for each action.

What Can You Actually Do with Jira MCP?

A lot! If you’re working in an Agile team, Jira MCP can simplify many of your daily workflows:

  • Building a new issue (e.g., bug, task, story)? Use the JIRA_CREATE_ISSUE tool to create a new ticket.
  • Want to comment on an issue? Use JIRA_ADD_COMMENT to drop a note.
  • Need to create tasks in bulk? The JIRA_BULK_CREATE_ISSUE tool has you covered.
  • Assign issues, fetch projects, list sprints, add attachments? You can do it all, by just passing simple prompts to your AI agent and watch it work.

Explore the full list of Jira MCP tools here

What’s Covered?

  • How to connect Composio’s Jira MCP and Claude Code in two different ways
  • How to configure Claude Code to manage Jira tickets from your terminal
  • How to prompt Claude to view, create, update, and delete Jira tasks

Let’s Start With the Setup

We’ll be using the following tools:

  1. Composio’s Jira MCP
  2. Claude Code
  3. Jira

You can set up Jira MCP in two ways:

Method 1: Quick MCP Setup (My Go-To)

Head over to Composio Jira MCP and hit the Claude tab. You’ll see a Generate button under Installation. Click it, and run the generated command in your terminal.

Composio MCP Dashboard

The command will look something like this:

npx @composio/mcp@latest setup "https://mcp.composio.dev/partner/composio/jira/mcp?customerId=[your-customer-id]" "jira-lig0gc-38" --client claude

You can follow the same instructions to set up the MCP for your Jira instance on Cursor, Windsurf, or even via a plain HTTP endpoint.

After running the command, the terminal will prompt you a success message:

Composio Success Message

By default, the config gets saved to:

  • C:\Users\[your-username]\AppData\Roaming\Claude\claude_desktop_config.json (on Windows)
  • ~/.config/claude/claude_desktop_config.json (on Mac/Linux)

Note: Before moving on, make sure you have the MCP config ready for the project you want to use it with. You can either use the global config or copy it into a local .mcp.json for individual projects. For this article, we’ll focus on a single project, so we’re going with a local config.

To copy the config file locally, run:

Windows:

cp C:\Users\[your-username]\AppData\Roaming\Claude\claude_desktop_config.json .mcp.json

Mac/Linux:

cp ~/.config/claude/claude_desktop_config.json .mcp.json

Configuration

MCP Connection

Now to connect MCP to Claude Code, you just need to authenticate it. Simply ask Claude to authenticate MCP. It’ll give you a URL to complete the OAuth flow. Once done, you’ll see a success message.

MCP Authentication

Boom, now you’re ready to start using MCP inside Claude Code.

Method 2: Dashboard Custom Server Setup

If you prefer doing things manually or want more control, here’s how to go about it:

  1. Log in to the Composio Dashboard
  2. On the All Apps page, search for Jira, select it, and hit Setup Jira IntegrationNew Integration.
  3. Head to the Atlassian Developer Console and create a new OAuth 2.0 integration.
  4. Name your app and add these scopes:
    Classic scopes:

    • read:jira-work
    • write:jira-work
    • manage:jira-project
    • manage:jira-data-provider

    Granular scopes:

    • read:board-scope:jira-software
    • write:board-scope:jira-software
    • read:project:jira
    • read:sprint:jira-software
    • write:sprint:jira-software and more if you want.
  5. Paste your Client ID and Secret into Composio Dashboard.

  6. Copy the Redirect URI from Composio and paste it into the OAuth setup on Atlassian.

  7. Set permissions under the Permissions tab.

Note: If granular scopes aren’t properly set, the authentication might fail and terminate the connection.

Once that’s done, go back to the Composio dashboard, finish the integration, and click Try Connecting Jira. Accept the permissions, and just like that, you’re connected.

Atlassian access

MCP Testing

Next, navigate to the tool page you just created, click Add MCP Server, name it, and run the following command:

npx @composio/mcp@latest setup "https://mcp.composio.dev/composio/server/[your-server-id]/mcp?connected_account_id=[your-connected-account-id]" "jira-mcp" --client claude

Note: You’ll find the server ID in the tool page URL, and the account-id will append only if you pass the account-id while creating the server.

After running the copy command (again, inside C:\\Users\\[your-username]\\AppData\\Roaming\\Claude on Windows or ~/.config/claude on Mac/Linux), your MCP server is ready to go with Claude.

Configured Servers

Managing Jira Tickets with Claude Code

Now that our config is ready, we can start managing Jira tickets from the Claude Code CLI.

I’ve already set up a project on Jira, here’s what it looks like with a few initial work items:

Jira Tickets

Navigate to your project directory (where you placed .mcp.json) and run Claude Code.

Prompt:

I need you to list all the tasks I have on my board 67 of Jira setup.

Claude Response

Done! You’ll get the list of tasks instantly. You can even ask Claude to:

  • Create a new task
  • Update an existing one
  • Delete tasks
  • …and more

A Video Walkthrough for the whole flow

Conclusion

Since I started using the MCP, I’ve been able to manage my projects way more efficiently, and finally spend more time on what actually matters: building stuff.

If you’re tired of the manual grind too, try this setup. It might just save your sanity.

Go to Composio MCP and BUILD NOWWWW!


This content originally appeared on DEV Community and was authored by Rohith Singh