Env Doctor
Diagnose and Fix CUDA / GPU environments compatibility issues locally, in Docker, and CI/CD. CLI + MCP server available.
What is Env Doctor?
Env Doctor is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to diagnose and fix cuda / gpu environments compatibility issues locally, in docker, and ci/cd. cli + mcp server available.
Diagnose and Fix CUDA / GPU environments compatibility issues locally, in Docker, and CI/CD. CLI + MCP server available.
This server falls under the Developer Tools category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Diagnose and Fix CUDA / GPU environments compatibility issue
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx env-doctorConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Env Doctor
Env Doctor is a Python CLI and MCP server that diagnoses and fixes GPU/CUDA environment compatibility issues for AI and machine learning workloads. It inspects the full GPU driver → CUDA toolkit → cuDNN → framework stack (PyTorch, TensorFlow, JAX) and generates safe, matched install commands, Dockerfile fixes, and GitHub Actions workflow templates. ML engineers use it to stop wasting hours debugging CUDA version mismatches and to verify whether a given model's VRAM requirements fit their hardware before downloading it.
Prerequisites
- Python 3.8+ with pip (or uv for isolated installation)
- A CUDA-capable GPU and NVIDIA driver installed (for GPU diagnostics; macOS supports non-GPU features only)
- Claude Desktop or another MCP-compatible client for AI assistant integration
- Optional: Docker installed if you want Dockerfile validation features
Install env-doctor
Install the core package with pip. For isolated, tool-style installation use uv. The optional `[dashboard]` extra adds a web UI for multi-machine fleet monitoring.
pip install env-doctor
# Or with uv (recommended for isolation):
uv tool install env-doctorRun a basic environment check
Verify the installation works and get your first environment diagnosis in the terminal.
env-doctor checkConfigure Claude Desktop
Add the env-doctor MCP server to your Claude Desktop config. The `env-doctor-mcp` command is installed by the package and exposes 11 diagnostic tools to AI assistants.
{
"mcpServers": {
"env-doctor": {
"command": "env-doctor-mcp"
}
}
}Restart Claude Desktop
Fully quit and reopen Claude Desktop so it registers the new MCP server. The env-doctor tools should then appear in the tools panel.
Check model VRAM requirements
Use the CLI to check whether a model fits in your GPU memory before downloading it. This calls the HuggingFace Hub to fetch model metadata.
env-doctor model meta-llama/Llama-3-70B-InstructGenerate a CUDA install sequence
Let env-doctor create a matched CUDA Toolkit installation plan. Use `--dry-run` to preview steps before executing, then `--run` to apply.
env-doctor cuda-install --dry-run
# When ready:
env-doctor cuda-install --runEnv Doctor Examples
Client configuration
Claude Desktop config connecting to the env-doctor MCP server using the installed entry point.
{
"mcpServers": {
"env-doctor": {
"command": "env-doctor-mcp"
}
}
}Prompts to try
After configuring Claude Desktop, use these prompts to diagnose and fix your GPU environment.
- "Check my GPU environment and tell me if anything is misconfigured."
- "Can I run Llama 3 70B on my GPU? How much VRAM does it need?"
- "Validate this Dockerfile for GPU issues: [paste Dockerfile contents]"
- "What PyTorch version should I install for my current CUDA driver?"
- "Generate a GitHub Actions workflow that validates my GPU environment in CI."
- "Install flash-attn safely for my current GPU and CUDA setup."Troubleshooting Env Doctor
'env-doctor-mcp' command not found when Claude Desktop tries to start the server
Ensure env-doctor is installed in the Python environment that Claude Desktop can access. If you used `uv tool install`, the binary is in `~/.local/bin/`; add that directory to your PATH or use the full path in the config command.
env-doctor reports no GPU found on a machine that has a GPU
On Linux, verify the NVIDIA driver is loaded with `nvidia-smi`. On Windows, ensure the driver is installed and not blocked by virtualization. On macOS, GPU diagnostics are not supported — only non-GPU features (model checks, Python compatibility, MCP server) are available.
CUDA install step fails partway through
Run `env-doctor cuda-install --dry-run` first to preview all steps. If a step fails during `--run`, check `/var/log/cuda-install.log` for errors and ensure you have sudo privileges. You can re-run from the failed step.
Frequently Asked Questions about Env Doctor
What is Env Doctor?
Env Doctor is a Model Context Protocol (MCP) server that diagnose and fix cuda / gpu environments compatibility issues locally, in docker, and ci/cd. cli + mcp server available. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Env Doctor?
Follow the installation instructions on the Env Doctor GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Env Doctor?
Env Doctor works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Env Doctor free to use?
Yes, Env Doctor is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Env Doctor Alternatives — Similar Developer Tools Servers
Looking for alternatives to Env Doctor? Here are other popular developer tools servers you can use with Claude, Cursor, and VS Code.
Ecc
★ 188.2kThe agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
Javaguide
★ 155.8kJava 面试 & 后端通用面试指南，覆盖计算机基础、数据库、分布式、高并发、系统设计与 AI 应用开发
Gemini CLI
★ 104.5kA secure MCP server that wraps the Google Gemini CLI, allowing clients to query Gemini models using local OAuth sessions without requiring an API key. It provides tools for model interaction and diagnostics with built-in protection against command in
Awesome MCP Servers
★ 87.3k⭐ Curated list of Model Context Protocol (MCP) servers - tools that extend Claude Desktop, Cursor, Windsurf, and other MCP clients with custom capabilities.
MCP Servers
★ 86.0kModel Context Protocol Servers
CC Switch
★ 77.5kA cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Gemini CLI & Hermes Agent. Only official website: ccswitch.io
Browse More Developer Tools MCP Servers
Explore all developer tools servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Env Doctor in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Env Doctor?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.