top of page
macOS


eSIM-Only at Fleet Scale: Test This Before the iPhone Duo Lands
The iPhone Duo is eSIM-only worldwide, but so is every iPhone 16 and 17 sold in the US. Here is what to test with your carrier and in Intune before October, and where the tooling is thinner than you would like.

MacSmithAI
2 days ago6 min read


Your Line-of-Business iOS Apps Are Not Ready for the iPhone Duo
The iPhone Duo's inner display reports regular size classes in both dimensions — the one assumption a decade of iPhone code takes for granted. Apple's developer guidance is out, the device ships 23 October, and here are the five questions to send your app owners now.

MacSmithAI
3 days ago6 min read


macOS Privacy Controls for AI Apps: TCC, Full Disk Access, and What to Allow
AI desktop apps want Accessibility, Screen Recording and sometimes Full Disk Access. Here's what a management profile can still grant, what it never could, and what changes when Apple removes the PPPC route for Accessibility in macOS 27.

MacSmithAI
5 days ago6 min read


Reviewing Shell Scripts with AI Before They Hit 10,000 Macs
A shell script pushed through Intune runs as root on every Mac in scope, and nothing in that pipeline reads your logic. Here is the three-layer review order that catches the mistakes — and exactly where AI review stops being trustworthy.

MacSmithAI
Sep 47 min read


Apple Put the M6 in a Mac mini First: What the Desktop Refresh Means for Your Fleet
Apple's 25 August desktop refresh debuts M6 silicon in the $899 Mac mini and pushes Mac Studio to M5 Max and M5 Ultra. Past the benchmark multipliers, here's what a Thunderbolt split, a 32GB ceiling and a 22 September ship date actually mean for a managed Mac fleet.

MacSmithAI
Sep 25 min read


PowerShell on macOS with AI: Graph Queries a Mac Admin Actually Runs
A bash-native Mac admin's setup for querying Intune device data with the Microsoft Graph PowerShell SDK — the install that actually belongs on a fleet, four queries worth keeping, and how to catch the cmdlets AI invents.

MacSmithAI
Sep 17 min read


Local LLMs on Apple Silicon: Ollama vs. LM Studio vs. MLX
The three-way framework debate mostly collapsed in 2026 — Apple's MLX is the engine underneath all of it now. Here is which wrapper to sanction on a managed Mac fleet, how much unified memory people actually need, and why port 11434 belongs in your firewall policy today.

MacSmithAI
Aug 316 min read


Writing Intune Compliance Scripts with Claude: A Bash Workflow That Actually Ships
Custom compliance settings for macOS just went GA in Intune, which means a lot of Mac admins are about to ask an AI to write their discovery scripts. Here is the prompt-review-test-deploy loop that keeps a fabricated preference key from locking out ten thousand users.

MacSmithAI
Aug 317 min read


The Mac Developer's AI Stack in 2026: VSCode, Raycast, Codex, Claude and More
The Mac has always been the developer's machine of choice. In 2026, it's also become the best platform for AI-assisted development — not because of any single tool, but because of how well a handful of them work together. Here's a practical breakdown of the AI stack serious Mac developers are running right now.

MacSmithAI
Jul 94 min read
bottom of page