VS Code with Stable Baseline.

Step 1 · Install in VS Code

Skip the config file. Click the button below and VS Code opens a confirmation panel with the server pre-filled. It will prompt you for your API key — paste your sta_… key once and it gets stored in the secure credential store, never on disk.

Step 2 · Create your API key

If you don't have an API key yet:

• Sign in to app.stablebaseline.io and open your project
• Open Project Settings (the gear icon)
• Navigate to MCP Setup → API Keys
• Click Create API Key, name it (e.g. “VS Code — Atlas”)
• Copy the sta_… key — it's shown once

Project-scoped vs organisation-wide keys

Project-scoped keys follow least-privilege — each dev project gets its own key, scoped to just that project's workspace. Full API key docs →

Or configure manually

Prefer editing files directly? Create a .vscode/mcp.json in your project root:

{
  "inputs": [
    {
      "type": "promptString",
      "id": "sb-api-key",
      "description": "Stable Baseline API Key",
      "password": true
    }
  ],
  "servers": {
    "sb": {
      "type": "http",
      "url": "https://api.stablebaseline.io/functions/v1/cloud-serve/mcp",
      "headers": {
        "Authorization": "Bearer ${input:sb-api-key}"
      }
    }
  }
}

Workspace vs user config

Step 3 · Verify the connection

• Open the Output panel (Ctrl+Shift+U / ⌘+Shift+U)
• Select MCP: sb from the dropdown (or GitHub Copilot Chat — MCP if using Copilot)
• Confirm Connection state: Running appears with a tool count in the 60+ range

Step 4 · Generate your documentation

Open GitHub Copilot Chat (Ctrl+Alt+I / ⌘+Alt+I) and run the setup prompt:

/mcp.sb.sb-setup

VS Code asks for prompt parameters (like project_goal) in an input bar at the top — you can type a short goal or press Enter for defaults. Pick either Insert as text or Run as Command and the agent begins onboarding your project.

What sb-setup does, in order

• Verifies the MCP connection — If it can't connect, it tells you exactly what to fix and waits for you to say “continue”
• Resolves your workspace & project — If you have multiple, it asks you to pick. If none, it links you to the app
• Creates .sb/config.json — Caches your workspace and project IDs locally
• Checks for existing documentation — If docs exist, asks whether to augment, replace, or cancel
• Analyses your repository — Tech stack, architecture, dependencies, security model, infrastructure, domain context
• Designs a bespoke documentation structure — No template. Tailored to your repo's natural boundaries
• Creates the documentation — Real content with diagrams, not stubs
• Generates AGENTS.md — File-pattern-to-document mappings specific to your codebase
• Provides a summary — Tree view of everything created, with next steps

How auto-sync works

The AGENTS.md generated in step 4 contains a table mapping your repo's file patterns to Stable Baseline documents:

These patterns are specific to your repo. When Copilot's agent changes files matching a pattern, it knows which doc to update — and does so automatically via the MCP tools.

File structure

your-project/
├── .vscode/
│   └── mcp.json          # MCP server config (step 1)
├── .sb/
│   └── config.json       # Project IDs (auto-generated)
├── AGENTS.md             # Cross-IDE auto-sync rules (auto-generated)
└── ...

All files are safe to commit. The API key is never stored in any config file — VS Code resolves it at runtime from its secure credential store. Add .sb/ to .gitignore if you prefer to keep IDs local.

Try it

Once setup has run, here are some prompts worth pasting into Copilot Chat:

Other prompts available