Atlassian Jira

You can connect Jira data to AI assistants through a dedicated MCP server, enabling natural language interactions to query projects, manage issues, and track work directly from your AI workflow. This setup keeps Jira data on your side while letting you ask questions like which issues are active in a project or to add comments, all without leaving your AI assistant.

How to use

Use an MCP-enabled AI assistant to access Jira data by wiring a dedicated MCP server into your assistant’s transport. You can ask questions, retrieve issue details, search across projects, create or update issues, and manage comments using natural language. For example, you can ask about active issues in a project, inspect issue details including comments, or add a new comment to an issue.

How to install

Prerequisites: ensure you have Node.js 18.0.0 or higher installed on your machine.

Install the MCP server globally so your AI assistant can connect to Jira.

npm install -g @aashari/mcp-server-atlassian-jira

Configuration and security

Configure the MCP server with your Atlassian Jira credentials. You will need your Jira site name, your user email, and an API token. Keep these credentials secure and do not share them.

For Claude Desktop, add the MCP server to your Claude configuration. The snippet below shows how to connect the Jira MCP server using npx with the required environment variables.

{
  "mcpServers": {
    "jira": {
      "command": "npx",
      "args": ["-y", "@aashari/mcp-server-atlassian-jira"],
      "env": {
        "ATLASSIAN_SITE_NAME": "your-company",
        "ATLASSIAN_USER_EMAIL": "your.email@company.com",
        "ATLASSIAN_API_TOKEN": "your_api_token"
      }
    }
  }
}

Troubleshooting

If you encounter authentication errors, verify that your API token has the correct permissions and that your site name matches your Jira URL. If you see 404 errors, double-check the API path casing and permissions. For problems with Claude Desktop integration, restart the client after updating the config and confirm the config file location on your OS.

Real-world usage notes

The MCP server exposes generic HTTP endpoint tools for Jira access (GET, POST, PUT, PATCH, DELETE) and supports JMESPath filtering to extract only the data you need. You can also toggle between TOON and JSON output formats depending on your needs for data transfer efficiency.

Tools and capabilities you can use through the MCP server

The MCP server exposes five generic tools to interact with Jira APIs: jira_get, jira_post, jira_put, jira_patch, and jira_delete. These allow you to query project data, create and update issues, manage comments, and perform other Jira REST API operations.

Examples of commands you can perform with the MCP server

You can list projects, fetch project details, retrieve issue information, search with JQL, create issues, add comments, update issues, and delete comments using the Jira REST API through MCP-enabled assistants.

Available tools

jira_get

GET any Jira API endpoint (read data) via MCP tools interface.

jira_post

POST to any Jira API endpoint (create resources) via MCP tools interface.

jira_put

PUT to any Jira API endpoint (replace resources) via MCP tools interface.

jira_patch

PATCH any Jira API endpoint (partial updates) via MCP tools interface.

jira_delete

DELETE any Jira API endpoint (remove resources) via MCP tools interface.