Convert Claude Code to Microsoft 365 with PowerShell

May 18 is the wall. After that, no new InfoPath forms can be published to SharePoint Online. July 14, the service goes dark entirely. Microsoft is not offering an extension. There is no automated migration tool. Every form becomes a manual rebuild in Power Apps, Power Automate, or Microsoft Forms. The frame everyone is using is “modernization.” The frame to actually use is “governance-cycle reset.” https://lnkd.in/eYuQVFWH

Here’s a tip for anyone who struggles with fragmented memory in Chat GPT and is still manually uploading documents. You can, of course, connect to a company SharePoint by going to Profile > Settings>Apps > Add More > Search “SharePoint.” But I have not found this to be entirely reliable. A fix that’s worked for me often is to first create a single parent folder in your shared drive (Google or SharePoint for instance) that houses every document you want to share with GPT. Make sure that the folder is shareable via URL.  In GPT, start all conversations pertinent to a specific topic (e.g.: SOP’s) within a Project ( I use this feature constantly). Then follow these steps inside the Project: Sources > Add Sources > Google Drive (or other MS shared drive) > Paste the aforementioned parent folder’s URL. In any conversation within that Project, GPT can now access all files saved within that parent folder. If it fails to locate a file, check the individual document’s security permissions to ensure it’s not restricted. To keep it organized and reduce search time, you may want to create multiple parent folders housing all files relevant to specific topics. For instance, create a Project relating entirely to SOP’s then link it to a parent folder in your shared drive that contains only SOP’s.

📢 𝙃𝙤𝙬 𝙩𝙤 𝙁𝙞𝙡𝙩𝙚𝙧 𝙎𝙪𝙗𝙜𝙧𝙞𝙙𝙨 𝙞𝙣 𝘿𝙮𝙣𝙖𝙢𝙞𝙘𝙨 365 𝙒𝙞𝙩𝙝𝙤𝙪𝙩 𝙈𝙤𝙙𝙞𝙛𝙮𝙞𝙣𝙜 𝙩𝙝𝙚 𝘿𝙖𝙩𝙖 𝙈𝙤𝙙𝙚𝙡 One of the common challenges in Microsoft Dynamics 365 projects is filtering subgrid data when there’s no direct relationship between the parent and child entities. The out-of-the-box features don’t always meet this requirement but with a bit of JavaScript, it’s entirely achievable. In this post, I’ll demonstrate how, on the Account form, changing the Industry field can dynamically filter the Contacts subgrid to display only those contacts that share the same industry without relying on any direct relationship between the two entities. 𝙇𝙞𝙣𝙠 𝙏𝙤 𝘽𝙡𝙤𝙜: https://lnkd.in/eyk9yM4T

Exciting news for Power Platform admins and enterprise organizations! Microsoft has just released Dataverse Admin Skills (dv-admin) in Public Preview — and it's a game-changer for how businesses manage their Dataverse environments at scale. Imagine this: instead of manually clicking through the Power Platform Admin Center across 20, 50, or 100 environments — you simply type: "Enable auditing on all my developer environments" The agent handles the rest. It lists environments, confirms with you, executes in parallel, and gives you a summary. No scripts. No browser tabs. At Quantril Technologies, this aligns directly with how we help our clients — from financial services to investment management — govern their CRM and Power Platform ecosystems with precision and speed. Key capabilities now in preview: — Settings Management across 37 PPAC toggles (audit, MCP, Fabric, security & more) — Bulk Delete jobs with built-in safety guardrails — Long-Term Retention for compliance-ready archival — Works inside Claude Code & GitHub Copilot via natural language This is what Agentic Administration looks like — and it's just the beginning. If your organization is running multiple Dataverse environments and wants to modernize how you manage governance, we'd love to connect. Read the full blog: https://lnkd.in/ebCyGaqQ #PowerPlatform #Dataverse #MicrosoftCopilot #EnterpriseAI #QuantrilTechnologies #CRM #Dynamics365 #AgenticAI #DigitalTransformation

Built and shipped a Microsoft Graph MCP server — connects Claude to Microsoft 365 across Entra ID users, Outlook mail and calendar, Teams, Planner, and OneNote. 54 tools, local Python process, sits next to the Desk365 MCP I shipped earlier. Here's what I'm using it for so far: Morning planner check. First thing I do is ask Claude what's on my Planner. Sets the day without me having to open the app. New hire creation. The Desk365 MCP pulls the new hire request and details, the Graph MCP creates the Entra account and assigns basic licensing — with me reviewing and approving at each step. What used to be a checklist across multiple tabs is now one conversation. SOPs and training docs. With OneNote in the mix, I can have Claude turn the workflow I just walked through into a structured SOP page in our team notebook — useful for handing off new-hire training without re-documenting the same steps from scratch every time. Teams to ticket flow. The one I didn't expect to like as much as I do. When a ticketable request comes in via Teams, I tell Claude who requested it and have it read the thread. If there's a resolution in the chat already, Claude pulls the full conversation, drafts the documentation, logs the time entry, and creates the ticket — closing it out if it's already resolved. Solves the "I helped them but never got around to ticketing it" problem that's been quietly inflating my unticketed work for months. The whole thing is just Claude orchestrating between Desk365 and Microsoft Graph. Neither MCP does anything novel on its own — the value is in the seam between them. #IT #Microsoft365 #MCP #ClaudeAI #Automation

Microsoft Syntex series: Which Syntex model should you use? When I first started exploring Microsoft Syntex, I assumed one model could manage every document type. But the more I worked with different specialties in terms of document types, the more I realised — choosing the right Syntex model is what actually makes automation useful. A lot of organisations today are sitting on thousands of files across SharePoint libraries, Teams folders, proposal archives, and contract systems....

🚀 Build agents for Microsoft 365 Copilot? There's a repo for that. The Microsoft 365 Copilot pro-developer samples repository is a community-driven collection of 75+ code samples showing you how to extend Microsoft 365 Copilot with custom agents — and it's growing every week. Whether you're just getting started or looking for advanced patterns, you'll find samples covering: 🤖 Declarative Agents — from blog post helpers and career coaches to IT help-desk and sales assistants ⚡ Custom Engine Agents — built with Microsoft 365 Agents SDK, Semantic Kernel, and MCP 🔌 API Plugins — connect to Azure Functions, Microsoft Graph, ServiceNow, Salesforce, and more 🔐 Authentication patterns — OAuth, API key, and Microsoft Entra ID 🛠️ Multiple languages & tools — TypeScript, C#, Python, TypeSpec, Dev Proxy, and Copilot Studio Some highlights: 📋 Restaurant ordering agents (JS, C#, Python, and Dev Proxy versions) 🎮 Geo Locator Game and Xbox Game Master 🏥 HR & IT Helpdesk agents with ServiceNow integration 📊 Sales Genie with Salesforce CRM 🎓 Canvas LMS agents for students and teachers 🔍 SharePoint Agents Finder using Microsoft Graph Search Every sample includes a README with setup instructions, architecture details, and screenshots so you can get up and running quickly. Built by the community, for the community. 💜 👉 Explore the samples: https://lnkd.in/ekHkg79B Want to contribute your own? Check out the contribution guide and share what you've built! #Microsoft365 #Copilot #AI #Agents #Microsoft365Dev #OpenSource #CopilotExtensibility #DeclarativeAgents #ProDev

🚀 Exploring the Power of Microsoft 365 Graph API If you're working with Microsoft 365, the Graph API is a game changer. It provides a unified endpoint to access data across services like Outlook, Teams, SharePoint, OneDrive, and more — all through a single REST API. 💡 Why it matters: * Centralized access to user, mail, calendar, and organizational data * Enables powerful automation and integrations * Supports real-time insights and intelligent workflows 🔧 What you can build: * Automated email and calendar tools * Teams bots and productivity apps * Data dashboards using organizational insights * Seamless integrations with third-party apps 📌 Key highlights: * Uses OAuth 2.0 for secure authentication * Rich SDK support (JavaScript, Python, .NET, etc.) * Continuously evolving with AI-powered capabilities If you're a developer looking to build smarter workplace solutions, Microsoft Graph is definitely worth diving into. #Microsoft365 #GraphAPI #APIs #CloudComputing #Developers #Automation #Productivity

"Where should I start learning Power Apps and Power Automate?" Lately, a lot of connections have reached out asking for the best resources to master the basics of the Power Platform. While there are endless tutorials out there, here is the exact roadmap I always recommend to build a solid foundation: 1.Start with the Source: Microsoft Learn Before jumping into video tutorials, bookmark the official Microsoft Learn Platform . It is completely free, systematically structured, and updated directly by the product teams. It gives you the "why" behind the architecture, governance, and data modeling. 2.For Deep-Dives & Real-World Use Cases: Reza Dorrani Once you have the basics down, head over to Reza Dorrani's YouTube Channel https://lnkd.in/gAcE_Rrg . His structured courses, step-by-step deep dives, and focus on enterprise best practices (especially around Dataverse and SharePoint) are unmatched. He teaches you how to build apps that actually scale. 3. For Practical, Hands-On Troubleshooting: Shane Young If you want someone to break down complex concepts into plain English with a side of humor, Shane Young's YouTube Channel https://lnkd.in/gC8cYUxT is your go-to. His videos are incredible for practical execution, functions, and solving those specific "how do I fix this error?" moments. My Advice: Don't just watch—build along. Pick a small manual process you do weekly and try to automate or build an app for it using these three resources. What resources helped you the most when you were starting out? Drop them in the comments! Best Regards, Gandla Veeresh, Mob:7981216570. #PowerPlatform #PowerApps #PowerAutomate #LowCode #MicrosoftLearn #ContinuousLearning

An automation pattern I keep coming back to: when a process creates recurring operational overhead, the fix usually isn't always writing more code. It's removing the manual steps people take before the automation even runs. Here's a version of it I've worked through. The task: schedule Azure Logic Apps to stop and start automatically. A solution already existed: a Python script using the Azure API. It worked well technically, but operationally it depended on server access, manual setup steps, scheduled task configuration, and infrastructure running continuously for a lightweight job. The solution I implemented: rebuilding the workflow with Power Automate connected to Azure, with a simple SharePoint interface on top. Now scheduling takes two steps. No dedicated server. No task scheduler setup. Less operational overhead. And importantly, it's accessible to non-technical team members who shouldn't need infrastructure knowledge just to schedule an automation. Same outcome. Far less friction. The Python solution was strong engineering. But "better" isn't always the most technically sophisticated approach. Sometimes it's the one people can use independently without support. Created the diagram using Eraser : https://www.eraser.io/ #PowerAutomate #Azure #Automation #LowCode #ProcessImprovement