A hands-on guide to the new Agents and Workflows in Copilot Studio. The retro-future game store, run by agents.
Copilot Studio has been rebuilt for complex, multi-step work, moving from a single chatbot to coordinated, multi-agent systems. This is a technical guide from the CAT team on building enterprise-grade scenarios on that new foundation: practical patterns for connected agents, MCP servers, runtime skills, and generated files, each one backed by a complete, runnable reference you can deploy and inspect. Copilot Studio has been rebuilt for complex, multi-step work, moving from a single chatbot to coordinated, multi-agent systems. BlastBox Omega is a complete, runnable reference: a parent agent that delegates to connected agents, calls multiple MCP servers, runs per-agent skills that generate and execute Python at runtime, and returns real generated files. End-to-end reference scenarios spanning agents and workflows in the new Copilot Studio.
What it shows
Every scenario is built from the same core capabilities of the Copilot Studio modern agent experience. The more advanced scenarios simply bring more of them together at once.
A parent orchestrator delegates to specialist child agents and weaves their answers back into a single conversation.
Reusable procedures the agent follows to get a task done, and they can reference assets like scripts and templates along the way.
Agents generate and run code at runtime to compute, transform, and analyze, going beyond text to real calculation.
Agents produce real deliverables, PDFs, images, and documents, that users can download and hand off.
The agent plans dynamically across turns, asks for clarification when a step is ambiguous, and revises its plan when the request changes.
Context and earlier results persist across the conversation, so the agent builds on what is already known instead of starting over.
Grounded answers from connected sources, documents, sites, and structured data.
Agents call external systems and APIs as tools through MCP servers and connectors, extending what they can do beyond built-in capabilities.
Walkthroughs
End-to-end transcripts that put the building blocks to work, from a quick self-serve task to a full multi-agent flagship. More scenarios will land here over time.
A member lost their BlastPass card. The agent verifies their identity, reissues the card, and renders a fresh tier-colored membership card. Membership MCP + card skill.
A defective console trade-up: rule the warranty, check stock, prorate the refund, apply loyalty points, and settle the balance. Two connected agents, four MCPs, chained skills, settlement PDF.
A manager works out end-of-quarter markdowns by drilling from a store-wide view down to one SKU week-by-week, then to policy-grounded discounts. Three connected agents, live charts, recommendations, manager PDF.
Get started
Import the solution, create the MCP connections, add the policy documents, publish the agents, then run the scripted prompts. Everything runs inside Power Platform. No servers, no code to host.
Upload the managed zip in Power Apps
Deploys the inline connector code to APIM
Click Create for each of the four connectors
Publish all four agents, then run the two scenarios
Go deeper
Reference material to help you understand and extend the new Copilot Studio agent experience, a technical deep dive on modern agents and the open-source plugin that ties into it.
A slide-by-slide walkthrough of the architecture behind the new Copilot Studio agents, connected agents, MCP, skills, orchestration, and file generation.
An open-source plugin for AI coding agents that migrates classic Copilot Studio agents to the modern experience, or builds brand-new modern agents from scratch.
View on GitHub