Jira MCP Server

This is a Model Context Protocol (MCP) server that exposes Jira API calls as tools for use with Claude and other AI models.

Setup

1. Install Dependencies

pip install -r requirements.txt

2. Configure Environment Variables

Create a .env file in the project root:

JIRA_EMAIL=your.email@gmail.com
JIRA_API_TOKEN=your_api_token_here
JIRA_BASE_URL=https://your-instance.atlassian.net

Note: Get your API token from: https://id.atlassian.com/manage-profile/security/api-tokens

3. Run the Client

python test_client.py

Available Tools

The MCP server exposes the following tools:

1. list_projects

List all Jira projects

• No parameters required

2. get_boards

Get all Agile boards

• No parameters required

3. get_board_columns

Get columns for a specific board

• board_id (integer): The ID of the board

4. get_issues_for_board

Get all issues for a specific board

• board_id (integer): The ID of the board

5. get_epics_from_board

Get all epics from a specific board

• board_id (integer): The ID of the board

6. create_issue

Create a new issue in Jira

• project_key (string): Project key (e.g., 'KAN')
• summary (string): Issue summary/title
• description (string): Issue description
• issue_type (string, optional): Issue type (Task, Epic, Story, etc.) - defaults to "Task"

7. delete_issue

Delete an issue from Jira

• issue_key (string): Issue key (e.g., 'KAN-4')

8. update_issue_status

Update the status of an issue

• issue_key (string): Issue key (e.g., 'KAN-1')
• transition_id (string): Transition ID for the new status

Usage with Claude

To use this MCP server with Claude, add it to your Claude config:

macOS/Linux: ~/.config/Claude/claude_desktop_config.json

Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "jira": {
      "command": "python",
      "args": ["/path/to/mcp_server.py"],
      "env": {
        "JIRA_EMAIL": "your.email@gmail.com",
        "JIRA_API_TOKEN": "your_api_token",
        "JIRA_BASE_URL": "https://your-instance.atlassian.net"
      }
    }
  }
}

Project Structure

.
├── test_client.py         # Client to run the server tools calls test
├── mcp_server.py          # Main MCP server implementation
├── api_calls/             # Original API call scripts (reference)
│   ├── create_issue_from_json.py
│   ├── delete_issue.py
│   ├── get_board_columns.py
│   ├── get_boards.py
│   ├── get_epics_from_board.py
│   ├── get_issues_for_board.py
│   ├── list_projects.py
│   └── update_issue_status.py
├── package.json           # Project metadata
├── requirements.txt       # Python dependencies
└── README.md             # This file

Key Improvements Over Individual Scripts

✅ Unified Interface: All API calls accessible via a single MCP server
✅ Security: Credentials managed via environment variables
✅ Error Handling: Consistent error handling across all tools
✅ Type Safety: Proper input schema validation
✅ Extensibility: Easy to add new tools
✅ Integration: Seamless integration with Claude and other AI models