Every week, I filter through everything Microsoft ships on Azure Logic Apps, AI Agents, API Management, Copilot Studio, and the broader Integration Platform — and surface only what working professionals actually need to know and act on. No press releases. No marketing fluff. Just the signal.

This week’s Signal – Gautam’s Take

The MCP governance gap is closing — and this week’s three announcements together close it completely.

Three things landed this week that fit a single pattern. First: the Azure API Center Plugin Marketplace entered public preview, giving teams a governed, RBAC-controlled endpoint where developers can discover and install approved MCP servers directly from Claude Code or GitHub Copilot CLI. Second: OneLake MCP went GA with 19 commands covering Fabric lakehouses, KQL databases, mirrored databases, and semantic models — completing the Microsoft data estate MCP coverage that began with Grafana MCP (operational telemetry) last week and Logic Apps MCP servers (enterprise workflows) the week before. Third: Azure Event Grid added Stripe as a native event source in public preview, meaning payment events (charges, subscriptions, invoice failures) can now flow directly into Azure’s event backbone without webhook infrastructure or custom brokers.

For integration practitioners, each of these reduces a category of custom integration work. The API Center marketplace replaces the wiki page of approved MCP servers with a governed catalogue your developers actually use. OneLake MCP means an agent can now traverse Fabric data alongside operational telemetry and enterprise workflows, all through the same MCP protocol, all authenticated via your existing Azure RBAC. The Stripe/Event Grid preview replaces a brittle custom webhook handler with a fully-managed event routing path to Azure Functions, Logic Apps, or Service Bus — the same destinations your other event-driven integrations already use.

AI + Integration

MCP Discovery & Governance

The Azure API Center Plugin Marketplace solves a problem that every team building MCP-connected agents hits around the fourth week: where do you put the list of approved MCP servers? The answer was previously ‘a wiki page and some hope.’ The marketplace endpoint gives you a governed, versioned, RBAC-controlled catalogue that integrates with the tools developers are already using.

How it works: enable the plugin marketplace on your API Center instance. A marketplace.git endpoint is provisioned at your data plane URL. Developers configure their GitHub Copilot CLI or Claude Code environment to point at this endpoint. They see your curated catalogue — not the entire internet. Register MCP servers, Skills, and AI plugins centrally, control who can discover what, and remove the coordination overhead from every team that needs to onboard an AI plugin.

• Requires: API Center instance + portal enabled → marketplace endpoint auto-provisioned
• Integrates with: GitHub Copilot CLI, Claude Code development environment
• What you register: MCP servers, Skills, AI plugins — same inventory as your API/agent registry
• Governance: existing API Center RBAC controls apply to plugin discovery

Fabric Data Bridge

With OneLake MCP GA, the MCP data coverage across Azure’s data estate is now substantial. Your agents have governed, authenticated access to:

• Operational telemetry — Azure Monitor, Application Insights, Kusto via Azure Managed Grafana MCP (GA, last week)
• Enterprise workflows — 1,400+ connectors via Logic Apps as MCP servers (preview, rolling)
• Analytical data — Fabric lakehouses, KQL, semantic models via OneLake MCP (GA this week)

All three use your existing Azure RBAC and managed identities. No separate auth stack. No custom MCP infrastructure. The pattern is: build your agent logic in Copilot Studio, expose your business workflows through Logic Apps MCP servers governed by APIM, let the agent query data through OneLake MCP and Grafana MCP, and govern discovery through API Center. That’s the full architecture.

Microsoft Announcements and Updates

• Cloud Cost Optimization: Principles that still matter
• Give your AI agent the keys to OneLake: OneLake MCP (Generally Available)
• Powering Event Driven Payments with Stripe and Azure Event Grid

Community Blog Posts

• 8 Tips for Writing Agent Skills by Philipp
• Chrome Skills: Getting Started with Google Cloud examples by Robin
• Giving OpenClaw a Brain Using Your Existing Azure Subscription by camerondwyer
• What are AI gateways in 2026, and do you actually need one now? by Sam
• Salesforce launches Headless 360 to turn its entire platform into infrastructure for AI agents by Michael 
• OpenAI takes aim at Anthropic with beefed-up Codex that gives it more power over your desktop by Zack
• AI Is Writing Our Code Faster Than We Can Verify It by Andrew 
• Are AI certifications worth the investment? by Paul
• Build to Learn vs Build to Earn by Marty
• Reluctantly Influential: Inside Lenny Rachitsky’s Demandingly Chill Life by Nate Martins
• The Evolution of My Coding Journey: From Pascal to AI by christian

Podcasts

• Context Is Everything: Getting the Most from GitHub Copilot with Joydip Kanjilal by The Modern .NET Show
• How AI Changes Development with Rob Conery by dotnetrocks

Video

• Introducing SQL projects in SSMS: A new Database DevOps workflow | Data Exposed by Microsoft
• Microsoft Agent Framework releasing version 1.0

Microsoft Learn Paths

• Set Up Standard Workflows as MCP Servers — Azure Logic Apps The official Microsoft Learn documentation for configuring Logic Apps Standard as an MCP server
• AI Gateway Capabilities in Azure API Management Comprehensive reference for all GenAI Gateway capabilities now available in Azure API Management — token quotas, semantic caching, circuit breakers, backend load balancing, and content safety integration.
• Migration approaches for BizTalk Server to Azure Logic Apps Migration approaches for BizTalk Server to Azure Logic Apps — Microsoft Learn reference covering assessment, pattern mapping, and toolchain selection including the new ODXtoWFMigrator.

That’s it for this week. If you found this useful, consider forwarding it to a colleague who works with Microsoft Integration Stack.

Have a tip, use case, or tool worth sharing? Reply to this email — I read every one.

Until next week,

Gautam