RAPP Brainstem Quick Start

The local-first agent server from the AIBAST Agents Library, powered by GitHub Copilot.

Agents Library · Production Guide · The AIBAST RAPP pattern

Recommended: VS Code with GitHub Copilot. Not required, but it's the best way to explore the AI frontier: Copilot installs the Brainstem, opens it in VS Code's browser and guides you through it. Get the prompt. Advanced: the one-liner below is a fully supported alternative that needs no VS Code.

macOS/Linux
# Works everywhere. Installs everything.
$ curl -fsSL https://microsoft.github.io/aibast-agents-library/install.sh | bash ⧉
Works on macOS & Linux. Auto-installs Python 3.11 if missing.
1

The Brainstem

Local agent server — the survival basics
Start here ▶

The biological brainstem keeps you alive — breathing, heartbeat, reflexes. Your AI brainstem does the same: it runs the core agent loop, calls tools, and responds. Everything else builds on top of this.

GitHub Copilot Chat in VS Code Agent
Get https://microsoft.github.io/aibast-agents-library/skill.md and do Tier 1 (The Brainstem) for me. Do every step yourself; only stop when I need to sign in.
Open Copilot Chat (Ctrl+Alt+I, Mac ⌃⌘I), choose Agent, paste, press Enter. Copilot installs it, waits for it to answer, opens it and gives you a tour.

Or run the one line yourself in a terminal:

The one-liner

This is the only step. It installs Python, clones the repo, sets up a venv, authenticates, and launches.

curl -fsSL https://microsoft.github.io/aibast-agents-library/install.sh | bash

Windows: irm https://raw.githubusercontent.com/microsoft/aibast-agents-library/main/install.ps1 | iex

Advanced — manual steps

1. Authenticate with GitHub

Your GitHub Copilot subscription IS the AI engine. No OpenAI keys needed.

gh auth login

2. Start the brainstem

The server starts on port 7071 with a chat UI, health endpoint, and agent auto-discovery.

brainstem

3. Talk to it

Open localhost:7071 in your browser, or hit the API directly.

curl -s localhost:7071/health | python3 -m json.tool

4. Make it yours

Write a soul.md to define its personality. Drop *_agent.py files into the agents folder to give it skills. Point SOUL_PATH and AGENTS_PATH in .env to your private files.

2

The Hippocampus

Persistent memory — runs local, deploys to Azure
Next step ▶
GitHub Copilot Chat in VS Code Agent
Get https://microsoft.github.io/aibast-agents-library/skill.md and do Tier 2 (The Hippocampus) for me. Do every step yourself; only stop when I need to sign in.
Open Copilot Chat (Ctrl+Alt+I, Mac ⌃⌘I), choose Agent, paste, press Enter. Copilot checks your Brainstem, signs you in to Azure in your own browser, and deploys.

Or do it yourself:

The hippocampus is where memories form. One command sets up everything — Python, Azure Functions Core Tools, venv, dependencies, start script. No Azure account or API keys needed. Auth happens through the chat UI.

Run the one-liner

This is the only step. It clones CommunityRAPP, installs everything, and tells you what to do next. The Microsoft-hosted setup guide documents the upstream dependency.

curl -fsSL https://raw.githubusercontent.com/microsoft/aibast-agents-library/main/community_rapp/install.sh | bash

Windows: irm https://raw.githubusercontent.com/microsoft/aibast-agents-library/main/community_rapp/install.ps1 | iex

Hippocampus setup and API guide

▶ Advanced: manual steps

Start it & authenticate

Open index.html in your browser and send a message. The chat UI walks you through GitHub device-code auth — no env vars to set.

cd ~/rapp-projects/my-project && ./start.sh

Deploy to Azure (when ready)

Edit local.settings.json with your Azure OpenAI credentials, then publish.

func azure functionapp publish YOUR_APP_NAME --build remote

Or deploy from the Azure Portal:

Deploy to Azure

Test the cloud deployment

curl -s https://YOUR_APP_NAME.azurewebsites.net/api/health | python3 -m json.tool
3

The Nervous System

Connect to Copilot Studio — reach into M365
Advanced ▶
GitHub Copilot Chat in VS Code Agent
Get https://microsoft.github.io/aibast-agents-library/skill.md and do Tier 3 (The Nervous System) for me. Do every step yourself; only stop when I need to sign in.
Open Copilot Chat (Ctrl+Alt+I, Mac ⌃⌘I), choose Agent, paste, press Enter. Copilot imports the Copilot Studio solution into the environment you pick and wires it to your Azure Function.

Or do it yourself:

Your nervous system extends your brain's reach to every part of your body — eyes, ears, hands. Copilot Studio does the same: it connects your agent to Teams, M365 Copilot, and the entire Microsoft ecosystem. Your brainstem becomes an enterprise AI.

1. Download the solution

Grab the Power Platform solution package from the repo. This .zip contains the declarative agent, connectors, and flows preconfigured for your Azure Function.

curl -fsSL -o MSFTAIBASMultiAgentCopilot.zip https://raw.githubusercontent.com/microsoft/aibast-agents-library/main/MSFTAIBASMultiAgentCopilot_1_0_0_5.zip

2. Import into Power Platform

Go to make.powerapps.com → Solutions → Import → upload the .zip. This creates the Copilot Studio agent and Power Automate flows in your environment.

3. Wire the connector

Open the imported Power Automate flow and update the HTTP action URL to point to your deployed Azure Function endpoint from Step 2. Add your Function Key if using key-based auth.

4. Publish to Teams & M365 Copilot

In Copilot Studio, publish the agent. Your users can now talk to it in Teams, M365 Copilot, and any channel Copilot Studio supports. Same agent logic you tested locally — now enterprise-wide.

What you learn: Copilot Studio, declarative agents, Power Platform solutions, Teams integration, enterprise AI deployment.

Checking brainstem...
Agents Library Production Guide GitHub Get help Hippocampus Installer Source Hippocampus Guide Brainstem Docs GitHub Models

Current Limitations

IMPORTANT: This is an experimental project managed by a v-team from the Artificial Intelligence Business Applications Specialist Team (AIBAST), not an officially supported Microsoft product.

This project is still in beta. Even if we're optimizing it for best-practice adherence, by using it you might sometimes experience unwanted patterns, errors, or simply bad architectures. By telling us through the Get help form you will help us improve these tools.

This project has adopted the Microsoft Open Source Code of Conduct. This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. MIT License — Copyright (c) 2026 Microsoft.