Jira MCP Server

Problems This Fork Resolves

1. Token-Efficient Sprint Issue Listing

When working with large sprints containing many issues, the standard list_issues_from_sprint tool returns full JSON responses that consume significant tokens, especially when LLMs need to analyze multiple sprints or boards. This fork adds a list_issues_from_sprint_csv tool that returns data in a compact CSV format, dramatically reducing token usage while preserving essential information (issue key, summary, status, assignee, estimates, fix versions, and hotfix labels).

2. Sprint State Filtering

The original list_sprints_from_board tool returns all sprints (active, closed, and future) without the ability to filter. This fork adds an optional state parameter that allows you to fetch only active, closed, or future sprints, reducing unnecessary data transfer and making it easier to focus on relevant sprints.

Key benefits:

• Reduced token consumption for sprint issue queries via CSV format
• Faster response times due to smaller payloads
• More cost-effective for large-scale sprint analysis
• Targeted sprint retrieval by state (active/closed/future)
• Maintains all critical issue information in a structured format

---

A Model Context Protocol Server for Jira.

Provides integration with Jira through MCP, allowing LLMs to interact with it.

Jira REST Api Docs

Installation

Manual Installation

Note: Requires Node version to be 22.12.0 or above

1. Create or get Jira Personal Access Token: Guide

2. Add server config to Claude Desktop:

   • MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
   • Windows: Check this Guide

{
  "mcpServers": {
    "jira": {
      "command": "npx",
      "args": ["-y", "@parassolanki/jira-mcp-server@latest"],
      "env": {
        "JIRA_PERSONAL_ACCESS_TOKEN": "email@example.com:your_personal_jira_access_token",
        "JIRA_BASE_URL": "jira_base_url"
      }
    }
  }
}

For Windows:

{
  "mcpServers": {
    "jira": {
      "command": "cmd /c npx",
      "args": ["-y", "@parassolanki/jira-mcp-server@latest"],
      "env": {
        "JIRA_PERSONAL_ACCESS_TOKEN": "email@example.com:your_personal_jira_access_token",
        "JIRA_BASE_URL": "jira_base_url"
      }
    }
  }
}

Components

Tools

1. list_projects: List projects from Jira.

   • Required inputs:
     • query (optional string): A query string used to filter the returned projects.
     • maxResults (optional number, max: 100): The maximum number of results to return.
     • expand (optional string): Expand additional information in the response. (comma separated description, lead, issueTypes, url, projectKeys, permissions and insight).

2. list_boards: List boards from a project.

   • Required inputs:
     • projectKeyOrId (string): Key or Id of the project.
     • name (optional string): Name of the project.
     • maxResults (optional number, max: 100): The maximum number of results to return.
     • startAt (optional number): The starting index of the returned boards.
     • type (optional string): The type of boards. (can be one of scrum or kanban).

3. list_sprints_from_board: List sprints from a board. Can filter by state (active, closed, or future).

   • Required inputs:
     • boardId (string): The ID of the board.
     • state (optional string): Filter sprints by state (can be one of active, closed, or future).
     • maxResults (optional number, max: 100): The maximum number of results to return.
     • startAt (optional number): The starting index of the returned boards.

4. list_issues_from_sprint: List issues from a sprint.

   • Required inputs:
     • boardId (string): The ID of the board.
     • sprintId (string): The ID of the sprint.
     • maxResults (optional number, max: 100): The maximum number of results to return.
     • startAt (optional number): The starting index of the returned boards.
     • expand (optional string): Expand additional information in the response. (comma separated schema and names).

5. create_issue: Create an issue in Jira (Only supports Task issue type).

   • Required inputs:
     • projectKeyOrId (string): Key or Id of the project.
     • summary (string): The summary/title of the issue.
     • description (string): The description of the issue.

Usage examples

Some example prompts you can use to interact with Jira:

1. "Show me all Jira projects" → execute the list_projects tool to see all available projects.
2. "What Kanban boards exist in the DEV project?" → execute the list_boards tool with the DEV project key and type parameter set to "kanban".
3. "Show me all the sprints for board ID 123" → execute the list_sprints_from_board tool to see all sprints associated with board 123.
4. "Show me only the active sprint for board ID 123" → execute the list_sprints_from_board tool with state parameter set to "active" to see only active sprints.
5. "What issues are in sprint 456 on board 123?" → execute the list_issues_from_sprint tool to see all issues in sprint 456 on board 123.
6. "Show me the first 50 issues from the current sprint on the Marketing board" → first execute list_boards to find the Marketing board ID, then list_sprints_from_board with state="active" to find the active sprint, then list_issues_from_sprint with maxResults=50.

Development

1. Install dependencies:

pnpm install

2. Configure Github Access token in .env:

JIRA_PERSONAL_ACCESS_TOKEN=email@example.com:your_personal_jira_access_token
JIRA_BASE_URL=jira_base_url

3. Run locally with watch:

pnpm dev

4. Build the server:

pnpm build

5. Local debugging with inspector:

pnpm inspector

TODOS

• list_projects
• list_boards
• list_sprints_from_board
• list_issues_from_sprint
• get_issue_by_id_or_key
• create_issue (task issue type only)
• create_issue (story, epic, sub-task issue types)
• update_issue
• delete_an_issue
• archieve_an_issue
• list_comments_from_issue
• get_comment_from_issue_by_id
• create_comment_in_issue
• update_comment_of_issue
• delete_comment_of_issue
• list_subtasks_from_issue
• get_user_by_username_or_key