How to Add LLMtoMD to VS Code with GitHub Copilot (MCP Setup)

GitHub Copilot's agent mode in VS Code can call tools — including MCP servers. Connect LLMtoMD and Copilot gains a searchable knowledge base of your FRD, API references, and design docs, so it builds against your actual requirements instead of forgetting them mid-session.

Here's the setup.

What you need

• VS Code with GitHub Copilot (agent mode).
• A free LLMtoMD account with at least one document uploaded. Connecting is free on every plan.

Option A — one click

On the LLMtoMD integrations page, click Add to VS Code. VS Code opens with an install prompt — click Install, then approve the sign-in. Done, no JSON.

Option B — manual config

Open the Command Palette (Ctrl/Cmd+Shift+P) → MCP: Open User Configuration, and add:

{
  "servers": {
    "llmtomd": { "type": "http", "url": "https://mcp.llmtomd.com/mcp" }
  }
}

Enable it in the MCP view. The first run opens a browser to sign in to LLMtoMD — approve it. No API key to juggle; revoke anytime under Settings → Connect your AI tools in LLMtoMD.

Also using Claude Code inside VS Code? It keeps its own config — see the Claude Code guide.

Verify it works

In Copilot Chat (agent mode), ask:

List my LLMtoMD documents.

Then put it to work:

Using my LLMtoMD documents, what does the spec say about the onboarding flow?

Copilot retrieves the answer from your FRD and codes against it.

Why it matters

Copilot is great at generating code and, like every agent, forgetful about why. Giving it a queryable project memory keeps generated features consistent with the spec, survives long sessions where context would otherwise compact away, and cuts the tokens you'd waste re-pasting requirements. The full picture is in Give Your AI Coding Agent a Memory.

---

Make your docs Copilot-ready. Convert your first document free →, or browse every integration.