A local-first AI agent server powered by GitHub Copilot.
No API keys. No cloud setup. Just your GitHub account.
Your GitHub Copilot subscription IS the AI engine. No OpenAI keys needed.
The server starts on port 7071 with a chat UI, health endpoint, and agent auto-discovery.
Open localhost:7071 in your browser, or hit the API directly.
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.
One script creates everything: Function App, Azure OpenAI, Storage Account, Application Insights. All using Entra ID auth โ no API keys.
Your local brainstem can now use Azure OpenAI and Azure Storage instead of GitHub Copilot. Update .env with your Azure endpoint, or keep running locally with Copilot โ your choice.
What you learn: ARM templates, Azure Functions (Python), managed identity, RBAC, Azure OpenAI.
The repo includes a Power Platform solution (MSFTAIBASMultiAgentCopilot_*.zip) that creates a declarative agent in Copilot Studio wired to your Azure Function.
Point the Copilot Studio action to your deployed Azure Function endpoint. The agent you built locally now answers in Teams and M365 Copilot.
Publish to your organization. Your brainstem โ the same agent logic you tested locally โ is now available to every user in your tenant.
What you learn: Copilot Studio, declarative agents, Power Platform solutions, Teams integration, enterprise AI deployment.