Bernhard the Builder
AboutAppsBlogNews

How to Install GitHub MCP in Antigravity without Docker

Here is how to run the GitHub MCP server locally with npx - without the annoying Docker error.

How to Install GitHub MCP in Antigravity without Docker
Quick Fix
Bernhard The Builder
2026-01-06
2 MIN READ
Download Resources
AI Directive

Install GitHub MCP Without Docker: When the Default Config Fails

I opened Antigravity today, ready to code, and ready to install another MCP. This time, GitHub MCP. And when I tried to install it, I was slapped in the face with this error:
bash
1exec: "docker": executable file not found in %PATH%
Are you kidding me?
I just want my AI agent to read a repository. I am not launching a microservice architecture. I am not orchestrating a Kubernetes cluster. I am reading text files from GitHub.
Why on earth does the default configuration assume I have a 4GB Docker Daemon running in the background just to fetch a JSON response?
>
The "Containerize Everything" mindset has gone too far. If it's a Node.js script, just let me run the Node.js script. Stop wrapping everything in virtualized heavy metal.
If you are like me, and you refuse to install Docker Desktop just to satisfy a default config setting, here is the fix.
GitHub MCPNo DockerAntigravity
Tell Antigravity to run the server as a naked Node process using npx. It’s faster, cleaner, and it actually works.

Step 1: Get The Keys

Before fixing the engine, get the keys. The AI needs permission to read your repos.
  1. Go to GitHub Settings -> Developer settings.
  2. Click Personal access tokens -> Tokens (classic).
  3. Generate a new token.
  4. Scopes: Select repo (full control) and user (read user data).
  5. Copy that token. If you lose it, you have to do this again.

Step 2: Open The Config

Antigravity hides this file, but it's the brain of your agent operations.
  1. Open the Agent Window in Antigravity.
  2. Click the three dots (...) in the top right.
  3. Select MCP Servers.
  4. Click Manage MCP Servers.
This opens mcp_config.json. This file is the Single Source of Truth for your tools.
Antigravity MCP Config

Step 3: The Surgery

Look for the github-mcp-server block. By default, it looks like a Docker command. Eviscerate that and replace it with a clean npx command.
Delete the old block and paste this in:
json
1"github-mcp-server": {
2  "command": "npx",
3  "args": [
4    "-y",
5    "@modelcontextprotocol/server-github"
6  ],
7  "env": {
8    "GITHUB_PERSONAL_ACCESS_TOKEN": "ghp_YOUR_TOKEN_HERE"
9  }
10}
Syntax Check //
Make sure you replace ghp_YOUR_TOKEN_HERE with the actual token you copied in Step 1. And for the love of clean code, watch your commas. One missing comma in this JSON file and the whole thing fails silently.

Step 4: The Hard Restart

This is where most people get stuck.
Simply reloading the window is not enough. The MCP servers are spun up when the application process starts.
  1. Close Antigravity completely.
  2. Wait a second.
  3. Open it again.

The Result

No Docker container. No virtual machine. No "executable not found" errors.
Just a direct, lightweight connection between your Agent and your Codebase.
The way it should have been from the start.
Now get back to building.

Latest Blog Posts

Manifesto2026-05-02

The Rise of the Agentic Internet

The era of building website content is dead. The digital world just hasn't seen the body yet. I am moving to Full Agentic AI — and the implications will dismantle the current server-based software industry.

2026-02-12

LM Studio vs. Ollama

LM Studio runs Llama 4 Scout on local GPUs - but even 96GB VRAM has limits. Context length matters. Kilo Code bridges your IDE to local models. Here is what I learned.

Best Practices2026-02-08

Why You Must Run ESLint Before You Touch the "Cloud"

Running ESLint locally isn't optional - it's your first defense against broken Vercel deployments. I learned this the hard way when my code pushed to Git, triggered Vercel, and failed after 5 minutes of waiting. The fix? A 0.5-second local ESLint check that catches errors before they reach production. Here's why ESLint prevents deployment failures, code rot, and invisible performance bugs.

Achievement2026-02-08

Building a Neural Link Architecture: Zero Link Rot with AI-Powered Semantic Linking

I got absolutely fed up with broken internal links and manual link maintenance. The problem? Hardcoded links rot when slugs change. The solution? A neural link architecture that uses vector embeddings, hybrid ranking algorithms, and AI to automatically inject semantically relevant links at render-time. This system eliminates link rot, scales to thousands of articles, and ensures every link is contextually relevant. Here's how I built a semantic linker that treats websites as living knowledge graphs for AI citation systems.

A
B
C

This article is part of a Semantic Cluster. All links are managed by the Digital Architect AI.