Jira MCP Server
Enables creating Jira issues and listing issue types for a Jira Cloud project using Basic Auth credentials.
README
Jira MCP Server
MCP (Model Context Protocol) server for Jira Cloud API integration. Create issues and list issue types using Basic Auth (email + API token).
Features
- jira_create_issue – Create a new Jira issue with project, summary, and issue type
- jira_list_issue_types – List available issue types for a project
Prerequisites
- Node.js 24+
- Jira Cloud site
- Atlassian API token
Setup
-
Clone and install
cd jira_mcp npm install -
Configure environment
Copy
.env.exampleto.envand fill in your values:cp .env.example .envJIRA_BASE_URL=https://your-domain.atlassian.net JIRA_EMAIL=user@example.com JIRA_API_TOKEN=your_api_token_here -
Build
npm run build
Usage
Run the server
npm start
Or with env vars inline:
JIRA_BASE_URL=https://your-domain.atlassian.net JIRA_EMAIL=you@example.com JIRA_API_TOKEN=xxx npm start
The server runs via stdio and is intended to be used as a subprocess by MCP clients (e.g. Cursor, Claude Desktop).
Cursor configuration
Add to your Cursor MCP config (e.g. ~/.cursor/mcp.json):
{
"mcpServers": {
"jira": {
"command": "node",
"args": ["/path/to/jira_mcp/dist/index.js"],
"env": {
"JIRA_BASE_URL": "https://your-domain.atlassian.net",
"JIRA_EMAIL": "your-email@example.com",
"JIRA_API_TOKEN": "your-api-token"
}
}
}
}
Tools
jira_list_issue_types
List issue types (Bug, Task, Story, etc.) for a project.
| Parameter | Type | Required | Description |
|---|---|---|---|
| project_key | string | Yes | Jira project key (e.g. PROJ) |
| response_format | string | No | markdown (default) or json |
jira_create_issue
Create a new Jira issue.
| Parameter | Type | Required | Description |
|---|---|---|---|
| project_key | string | Yes | Jira project key |
| summary | string | Yes | Issue title |
| issue_type | string | Yes | Issue type name (e.g. Bug, Task) or ID |
| description | string | No | Issue description |
| response_format | string | No | markdown (default) or json |
Testing Without Cursor
The best way to test the MCP server interactively is with MCP Inspector, a web UI for exercising tools, resources, and prompts.
-
Build the project (if not already done):
npm run build -
Run MCP Inspector with your server. With a
.envfile in the project root, the server loads it automatically:npm run inspectorOr run directly (env from
.envis loaded by the server):npx @modelcontextprotocol/inspector node dist/index.jsTo pass env vars explicitly instead:
npx @modelcontextprotocol/inspector \ -e JIRA_BASE_URL=https://your-domain.atlassian.net \ -e JIRA_EMAIL=you@example.com \ -e JIRA_API_TOKEN=your-token \ node dist/index.js -
Open the Inspector in your browser (typically
http://localhost:6274). -
Test tools – Use the Tools tab to call
jira_list_issue_typesandjira_create_issuewith sample inputs.
The Inspector runs your server as a subprocess and connects via stdio, so it behaves like a real MCP client.
Development
# Run with auto-reload
npm run dev
# Run all tests (unit + integration)
npm test
# Run tests with coverage
npm run test:coverage
# Run integration tests only (spawns MCP server + mock Jira)
npm run test:integration
# Lint code
npm run lint
npm run lint:fix
# Format code
npm run format
npm run format:check
Pre-commit hooks (Husky)
On each git commit, Husky runs lint and Prettier on staged files. If there are errors, the commit is blocked with a message:
❌ Commit blocked: Fix the lint or Prettier errors above.
Tips: npm run lint:fix | npm run format
To bypass (use sparingly): git commit --no-verify or HUSKY=0 git commit
CI
GitHub Actions runs on pull requests to main/master with separate stages on Node.js 24.x:
- Unit Tests –
npm run test:unit(90% coverage enforced) - Integration Tests –
npm run test:integration(MCP server + mock Jira) - Lint & Format – ESLint and Prettier checks
- Security Audit –
npm audit(fails on moderate+ vulnerabilities) - CodeQL – Static security analysis (JavaScript/TypeScript)
- Dependency Review – Checks for vulnerable deps in PRs
Dependabot is configured for weekly dependency updates (.github/dependabot.yml).
To block merges until PRs pass, enable branch protection in GitHub: Settings → Branches → Add rule → Require status checks to pass before merging.
API Reference
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。