mcp-jira-python
MCP server for JIRA integration enabling issue management, search, comments, and linking via natural language.
README
MCP JIRA Python 🚀
A Python implementation of a MCP server for JIRA integration. MCP is a communication protocol designed to provide tools to your AI and keep your data secure (and local if you like). The server runs on the same computer as your AI application and the Claude Desktop is the first application to run MCP Servers (and is considered a client. See the examples folder for a simple python MCP client).
Installation
# Install the server locally
git clone https://github.com/kallows/mcp-jira-python.git
Tools Available
This MCP server provides the following JIRA integration tools:
delete_issue: Delete a Jira issue or subtask using its issue keycreate_jira_issue: Create a new Jira issue with customizable fields including summary, description, type, priority, and assigneeget_issue: Retrieve complete issue details including comments and attachments for a given issue keycreate_issue_link: Create relationships between issues (e.g., "blocks", "is blocked by", etc.)update_issue: Update existing issues with new values for fields like summary, description, status, priority, or assigneeget_user: Look up a user's account ID using their email addresslist_fields: Get a list of all available JIRA fields and their propertieslist_issue_types: Retrieve all available issue types in your JIRA instancelist_link_types: Get all possible relationship types for issue linkingsearch_issues: Search for issues using JQL (JIRA Query Language) within a specific projectadd_comment: Add a text comment to an existing issueadd_comment_with_attachment: Add a comment to an issue with an attached file
Claude Desktop Configuration
This requires you update claude_desktop_config.json. The file's location varies depending on Apple, Windows, or Linux.
Windows
Note: location of claude_desktop_config.json in Windows is:
%AppData%\\Claude\\claude_desktop_config.json
This will resolve (usually) to: C:\Users\YOURUSERNAME\AppData\Roaming\Claude
Below is the configuration block to add to claude_desktop_config.json.
With Windows we always use full paths. You will update "command", set your directory path, and add your JIRA env settings
<pre>
"jira-api": {
"command": "C:\\Users\\YOURUSERNAME\\.local\\bin\\uv.exe",
"args": [
"--directory",
"D:\\mcp\\mcp-jira-python",
"run",
"-m",
"mcp_jira_python.server"
],
"env": {
"JIRA_HOST": "YOURNAME.atlassian.net",
"JIRA_EMAIL": "yourname@example.com",
"JIRA_API_TOKEN": "YOURJIRATOKEN"
}
}
</pre>
☠️WARNING - you MUST close Claude Desktop AND kill all Claude processes to enable the updated claude_desktop_config.json!😬
Mac and Linux
Update the filepath to mcp-jira-python and fill in your JIRA env values:
<pre>
"mcp-jira-python": {
"command": "uv",
"args": [
"run",
"--directory", "/your/filepath/mcp-jira-python",
"src/jira-api/server.py"
],
"env": {
"JIRA_HOST": "your_org.atlassian.net",
"JIRA_EMAIL": "you@your_org.com",
"JIRA_API_TOKEN": "your_api_token"
}
}
</pre>
Note:
You must restart Claude Desktop after saving changes to claude_desktop_config.json.
Running MCP JIRA Python Tools
These MCP Tools are listed under jira-api server. You can see the listing by clicking on the tiny hammer in the lower right corner of the Claude Desktop text entry box. Please verify that the jira-api tools are available in the list. To 'run' a tool, just ask Claude specifically to do a Jira task. Notably, Claude may not see the tools at first and has to be nudged. In some cases, he will refuse to use tools. Updating the system prompt is recommended.
Running Tests
TODO - add description of running the tests (unittest) TODO - add some code to make it easier for the tests to get the env vars as the integration and system tests require the following environment variables:
export JIRA_HOST="your-domain.atlassian.net"
export JIRA_EMAIL="your-email@example.com"
export JIRA_API_TOKEN="your-api-token"
TODO - generate a test coverage report:
Project Structure
mcp-jira-python/
├── README.md
├── pyproject.toml
├── src/
│ └── jira_api/
│ ├── __init__.py
│ └── server.py
└── tests/
├── __init__.py
├── test_jira_api.py
├── test_jira_integration.py
└── test_jira_mcp_system.py
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。