Automating Engineering Workflows: How Miro uses MCP & AI Agents
In this webinar, the Miro team explores how the Model Context Protocol (MCP) allows models to move beyond answering questions to performing complex tasks like cooking up code, diagrams, and documentation directly within Miro. We walk through efficiency gains created by using our own MCP Server internally. We cover using it for security reviews, PR Reviews, logs tracing, and building slide decks for leadership presentations.
What you’ll learn:
- The Power of MCP: How it enables AI agents to read and write data across multiple tools seamlessly.
- Engineering Use Cases: Automating security reviews, visualizing legacy codebases, and converting technical designs back into code.
- Management Workflows: Using AI to surface system issues, build dependency graphs from logs, and auto-generate executive slide decks.
- Real-time Demos: Watch our team use Claude and Miro MCP to automate complex documentation and review processes in minutes.
Resources:
Official Miro MCP Docs: https://developers.miro.com/docs/mcp-intro
Open Source Repo (Plugins & Skills): https://github.com/miroapp/miro-ai
Chapters
00:00 Introduction to Model Context Protocol (MCP)
00:51 Miro’s MCP Server Strategy
01:45 Key Functionalities: Docs, Tables, & Diagrams
02:58 Primary Use Cases for Engineers
04:54 Demo: Automating Internal Security Reviews
08:22 Sample Prompt & Live Agent Execution
10:20 Reviewing AI-Generated Artifacts
13:30 Managing Complex PR Reviews with Visuals
16:11 Management Perspective: Problem Solving in One Place
18:21 Surfacing System Issues & Observability Gaps
20:57 Context to Slide Decks: Communicating Upwards
23:42 Summary & Developer Resources
#MiroMCP #MiroAI #aiagents