openroad-mcp
A Model Context Protocol (MCP) server that provides tools for interacting with OpenROAD and ORFS (OpenROAD Flow Scripts).
README
OpenROAD MCP Server
<!-- mcp-name: io.github.luarss/openroad-mcp -->
A Model Context Protocol (MCP) server that provides tools for interacting with OpenROAD and ORFS (OpenROAD Flow Scripts).
Demo

Features
- Interactive OpenROAD sessions - Execute commands in persistent OpenROAD sessions with PTY support
- Session management - Create, list, inspect, and terminate multiple sessions
- Command history - Access full command history for any session
- Performance metrics - Get comprehensive metrics across all sessions
- Report visualization - List and read report images from ORFS runs
Requirements
- OpenROAD installed and available in your PATH
- OpenROAD-flow-scripts (ORFS) for complete RTL-to-GDS flows (optional but recommended)
- Python 3.13+ or higher
- uv package manager
- Install:
curl -LsSf https://astral.sh/uv/install.sh | sh
- Install:
Support Matrix
| MCP Client | Supported | Transport Mode(s) | Notes |
|---|---|---|---|
| Claude Code | ✅ | STDIO | Full support for all features |
| Gemini CLI | ✅ | STDIO | Full support for all features |
| Other MCP clients | ⚠️ | STDIO | Should work with standard STDIO transport |
Getting Started
New to OpenROAD MCP? Check out our Quick Start guide.
For platform-specific setup instructions, see the Cross-Platform Guide.
Standard Configuration
The basic configuration for all MCP clients:
{
"mcpServers": {
"openroad-mcp": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/luarss/openroad-mcp",
"openroad-mcp"
]
}
}
}
For local development, use:
{
"mcpServers": {
"openroad-mcp": {
"command": "uv",
"args": [
"--directory",
"/path/to/openroad-mcp",
"run",
"openroad-mcp"
]
}
}
}
Installation
<details> <summary><b>Claude Code</b></summary>
Use the Claude Code CLI to add the OpenROAD MCP server:
claude mcp add --transport stdio openroad-mcp -- uvx --from git+https://github.com/luarss/openroad-mcp openroad-mcp
</details>
<details> <summary><b>Gemini CLI</b></summary>
Follow the Gemini MCP install guide, using the standard configuration above.
</details>
<details> <summary><b>Docker</b></summary>
🚧 Work in Progress: Docker deployment via GitHub Container Registry (GHCR) is coming soon.
</details>
<details> <summary><b>MCP Registry</b></summary>
Once published to the MCP Registry, clients can discover and install directly:
uvx openroad-mcp
</details>
Verification
After configuration, restart your MCP client and verify the MCP server is running:
- The server should automatically start when your MCP client launches
- You can use OpenROAD tools through the MCP interface
- Check logs for any startup errors if tools are not available
Available Tools
Once configured, the following tools are available:
interactive_openroad- Execute commands in an interactive OpenROAD sessioncreate_interactive_session- Create a new OpenROAD sessionlist_interactive_sessions- List all active sessionsterminate_interactive_session- Terminate a sessioninspect_interactive_session- Get detailed session informationget_session_history- View command historyget_session_metrics- Get performance metricslist_report_images- List ORFS report directory imagesread_report_image- Read a ORFS report image
Troubleshooting
If the MCP server fails to start:
- Ensure
uvis installed and available in your PATH - Verify the path to openroad-mcp is correct
- Check that all dependencies are installed:
make sync - Review your MCP client logs for specific error messages
Development
Setup
# Install environment
uv venv
make sync
Testing
# Run core tests (recommended - excludes PTY tests that may fail in some environments)
make test
# Run interactive PTY tests separately (may have file descriptor issues in CI)
make test-interactive
# Run all tests including potentially problematic PTY tests
make test-all
# Format and check code
make format
make check
Note: Interactive PTY tests are separated because they may experience file descriptor issues in certain environments (containers, CI systems). The core functionality tests (make test) provide comprehensive coverage of the MCP integration without these environment-specific issues.
MCP Inspector
# Launch MCP inspector for debugging
# For STDIO transport: Set Command as "uv", Arguments as "run openroad-mcp"
make inspect
Contributing
We welcome contributions to OpenROAD MCP! Please see CONTRIBUTING.md for detailed instructions on how to get started, our development workflow, and code standards.
Support
If you encounter any issues or have questions, please open an issue on our GitHub issue tracker.
License
BSD 3-Clause License. See LICENSE file.
Built with ❤️ by Precision Innovations
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。