mcp-pagure
Enables AI assistants to manage issues, pull requests, and projects on Pagure instances (pagure.io and self-hosted).
README
mcp-pagure
MCP server for Pagure - enables AI assistants to manage issues, pull requests, and projects on Pagure instances (pagure.io and self-hosted).
Features
- Read Operations: List/get projects, issues, pull requests, branches, and user info
- Write Operations: Create issues, add comments, update status, assign issues, merge/close PRs
- Multi-instance Support: Works with pagure.io and self-hosted Pagure instances
- Namespace Support: Full support for namespaced projects (rpms/, container/, etc.)
Installation
Using uvx (recommended)
uvx mcp-pagure
Using pip
pip install mcp-pagure
Configuration
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
PAGURE_API_TOKEN |
Yes | - | Your Pagure API token |
PAGURE_URL |
No | https://pagure.io |
Base URL for Pagure instance |
Getting an API Token
- Go to your Pagure instance (e.g., https://pagure.io)
- Navigate to Settings → API Keys
- Create a new token with the permissions you need:
issue_create- Create new issuesissue_update- Update issues (status, assignee, comments)pull_request_merge- Merge pull requestspull_request_close- Close pull requestspull_request_comment- Comment on pull requests
Usage with Claude Code
Add to your Claude Code MCP configuration (~/.claude/claude_code_config.json):
{
"mcpServers": {
"pagure": {
"command": "uvx",
"args": ["mcp-pagure"],
"env": {
"PAGURE_API_TOKEN": "your-api-token-here",
"PAGURE_URL": "https://pagure.io"
}
}
}
}
For Fedora's src.fedoraproject.org
{
"mcpServers": {
"fedora-dist-git": {
"command": "uvx",
"args": ["mcp-pagure"],
"env": {
"PAGURE_API_TOKEN": "your-fedora-token",
"PAGURE_URL": "https://src.fedoraproject.org"
}
}
}
}
Available Tools
Read Operations
| Tool | Description |
|---|---|
whoami |
Get information about the authenticated user |
list_projects |
Search/list projects with filters (pattern, owner, namespace, fork status) |
get_project |
Get detailed project information |
list_issues |
List issues with filtering (status, tags, assignee, author, search) |
get_issue |
Get issue details including all comments |
list_pull_requests |
List PRs with filtering (status, assignee, author) |
get_pull_request |
Get PR details including commits and comments |
get_user |
Get user profile information |
list_branches |
List git branches in a repository |
Write Operations
| Tool | Description |
|---|---|
create_issue |
Create a new issue with title, description, tags, assignee |
add_issue_comment |
Add a comment to an existing issue |
update_issue_status |
Change issue status (Open/Closed) with optional close reason |
assign_issue |
Assign or unassign an issue to a user |
add_pr_comment |
Add a comment to a pull request (general or inline) |
merge_pull_request |
Merge a pull request |
close_pull_request |
Close a pull request without merging |
Examples
List issues in a project
List open issues in the pagure project
Create an issue
Create an issue in mcp-pagure titled "Add webhook support" with description "We should add support for Pagure webhooks to enable real-time notifications"
Work with namespaced projects
List pull requests in rpms/kernel
Search for projects
Find projects matching "container"
Development
Setup
git clone https://github.com/fatherlinux/mcp-pagure.git
cd mcp-pagure
uv sync --dev
Run tests
uv run python -m pytest -v
Run the server locally
export PAGURE_API_TOKEN="your-token"
uv run mcp-pagure
Test with MCP Inspector
npx @modelcontextprotocol/inspector uv run mcp-pagure
Supported Pagure Instances
- pagure.io - Main Pagure instance
- src.fedoraproject.org - Fedora dist-git
- Any self-hosted Pagure instance
License
MIT License - see LICENSE for details.
Contributing
Contributions are welcome! Please feel free to submit a Pull Request.
Related Projects
- Pagure - The git-centered forge
- Model Context Protocol - The MCP specification
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。