How to Connect LLMtoMD to Google Antigravity (MCP Setup)
By The LLMtoMD team
Google Antigravity is an agent-first coding environment — and like every agent, it builds better when it can reference your actual requirements instead of guessing. Connecting LLMtoMD over MCP gives it a searchable knowledge base of your FRD, specs, and design docs.
Antigravity's MCP config has two small quirks worth getting right, so follow the steps below.
What you need
- Google Antigravity installed.
- A free LLMtoMD account with a document or two uploaded. Connecting is free on every plan.
- An LLMtoMD API key — generate one under Settings → Connect your AI tools.
Add the server
- In Antigravity: Settings → Customizations → Open MCP Config (or edit
~/.gemini/config/mcp_config.jsondirectly). - Add the entry below. Two things to note: Antigravity uses
serverUrl(noturl), and you authenticate with an API key in the Authorization header — its OAuth sign-in for custom servers is new and can hang mid-flow, so the key is the reliable path.
{
"mcpServers": {
"llmtomd": {
"serverUrl": "https://mcp.llmtomd.com/mcp",
"headers": { "Authorization": "Bearer mic_YOUR_KEY" }
}
}
}
- Save and restart Antigravity. The LLMtoMD tools appear under Manage MCP servers.
Verify it works
Ask the agent:
List my LLMtoMD documents.
Then put your spec to work:
Using my LLMtoMD documents, what are the requirements for the checkout flow?
It pulls the answer from your FRD and builds against it.
Why it matters
Antigravity's agent can run long, multi-step builds — exactly where context drifts and the original spec gets lost. A connected knowledge base keeps it grounded: it can search_documents and ask_documents whenever a requirement is unclear, so generated code stays on-spec. The full case (and the token savings) is in Give Your AI Coding Agent a Memory.
Give your agent the spec, not a guess. Convert your first document free →, or see every integration.
Convert anything to AI-ready Markdown
PDFs, Office docs, images, audio, and whole websites — clean Markdown and RAG-ready exports for your LLM, in seconds.