home / mcp / atlassian mcp server

Atlassian MCP Server

Jira MCP Server

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "amruthabesto28-jira-mcp": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "-v",
        "\"<path_to_your_home>/.mcp-atlassian:/home/app/.mcp-atlassian\"",
        "-e",
        "JIRA_URL",
        "-e",
        "ATLASSIAN_OAUTH_CLIENT_ID",
        "-e",
        "ATLASSIAN_OAUTH_CLIENT_SECRET",
        "-e",
        "ATLASSIAN_OAUTH_REDIRECT_URI",
        "-e",
        "ATLASSIAN_OAUTH_SCOPE",
        "-e",
        "ATLASSIAN_OAUTH_CLOUD_ID",
        "ghcr.io/sooperset/mcp-atlassian:latest"
      ],
      "env": {
        "JIRA_URL": "https://your-company.atlassian.net",
        "ATLASSIAN_OAUTH_SCOPE": "read:jira-work write:jira-work offline_access",
        "ATLASSIAN_OAUTH_CLOUD_ID": "YOUR_CLOUD_ID_FROM_SETUP_WIZARD",
        "ATLASSIAN_OAUTH_CLIENT_ID": "YOUR_OAUTH_APP_CLIENT_ID",
        "ATLASSIAN_OAUTH_REDIRECT_URI": "http://localhost:8080/callback",
        "ATLASSIAN_OAUTH_CLIENT_SECRET": "YOUR_OAUTH_APP_CLIENT_SECRET"
      }
    }
  }
}

You can connect Jira to an MCP (Model Context Protocol) server to automate updates, perform intelligent searches, and manage issues across Jira Cloud or Server/Data Center. This MCP acts as a bridge between your AI assistant and Jira, enabling tasks like creating issues from notes, filtering urgent items, and performing JQL searches with natural language prompts.

How to use

You interact with the Jira MCP server from your AI-enabled IDE or client by selecting Jira as a connected MCP server and then issuing high-level commands. Typical workflows include updating Jira from meeting notes, filtering for urgent issues, or locating issues assigned to you that are due soon. The MCP translates your requests into Jira API calls and returns structured results for display in your tooling.

If you are using OAuth-based authentication, ensure tokens are configured for each cloud instance you work with. If you operate in BYOT mode, you provide a token yourself and manage its lifecycle. You can also route API traffic through proxies or add custom HTTP headers to satisfy corporate network requirements.

How to install

Prerequisites: install Docker on your machine. You will run the MCP Jira server as a Docker container.

1) Pull the pre-built image for the MCP Jira server.

2) Run the container with your Jira configuration and authentication options. Use the examples below to start with standard OAuth flow or BYOT tokens.

OAuth 2.0 Configuration (Cloud) and BYOT details

The setup supports two OAuth-related approaches. Use the standard OAuth flow with the setup wizard or supply an existing OAuth access token directly (BYOT). Both methods are started via Docker, and you provide the necessary environment variables.

Security and advanced configuration

Configure authentication methods, HTTP proxies, and custom headers to align with your security and networking requirements. The server supports multi-cloud OAuth setups, token-based authentication, and per-service proxy overrides. Adjust logging and verbose modes to help diagnose issues without exposing sensitive data.

Available tools

jira_get_issue

Retrieve the details of a single Jira issue by its key.

jira_search

Query Jira issues using JQL and return matching results.

jira_create_issue

Create a new Jira issue with provided fields.

jira_update_issue

Update the fields of an existing Jira issue.

jira_transition_issue

Transition an issue to a new status.

jira_add_comment

Add a comment to a Jira issue.

jira_get_project_issues

List issues for a given Jira project.

jira_get_board_issues

List issues on a Jira board.

jira_get_sprints_from_board

Fetch sprints for a given board.

jira_get_user_profile

Retrieve user profile information.

jira_download_attachments

Download attachments from an issue.