17.04.2026
Daring Fireball
• John Gruber
WorkOS FGA: The Authorization Layer for AI Agents
My thanks to WorkOS for sponsoring last week at DF. Every AI agent demo looks magical, but most hit a wall in enterprise deployment. It’s not model quality or latency. It’s authorization. Authentication proves an agent’s identity. Authorization defines its blast radius.
The winners in enterprise AI won’t have the most features. They’ll be the ones enterprises can safely trust. Learn how WorkOS FGA scopes that blast radius with resource-level permissions, and read their deep dive for more.
Woohoo! Jasper and Florian added webmention support to their blogs!
Iran closes Strait of Hormuz a day after declaring it open
Iran says its forces are in full control of the Strait of Hormuz and passage through the vital oil waterway would remain restricted, unless the US removed its blockade of Iranian ports. President Trump has said the US will continue to stop ships going to or coming from Iran ...
Changes in the system prompt between Claude Opus 4.6 and 4.7
Anthropic are the only major AI lab to publish the system prompts for their user-facing chat systems. Their system prompt archive now dates all the way back to Claude 3 in July 2024 and it's always interesting to see how the system prompt evolves as they publish new models.
...
Claude system prompts as a git timeline
Research: Claude system prompts as a git timeline
Anthropic publish the system prompts for Claude chat and make that page available as Markdown. I had Claude Code turn that page into separate files for each model and model family with fake git commit dates to enable browsing the changes via the GitHub commit view.
Tags: system-prompts, anthropic, claude, generative-ai, ai, llms
The Happy Pod: Filling the empty chair left by my best friend
We meet the man helping to tackle loneliness by offering strangers somewhere to sit and chat. Dean Perryman started the Empty Chairs project, honouring the memory of his best friend. After he posted the concept online, it has been replicated around the world.
Also: How tech...
Daring Fireball
• John Gruber
★ ‘A Reading Room on Wheels, a Lover’s Lane, and, After 11 PM, a Flophouse’
Vittoria Benzine, at Artnet (via Oliver Thomas):
The singular American filmmaker Stanley Kubrick saw the little
details. He even saw the future. But, most of all, he saw
people, with all their quirks. Kubrick’s films, from Dr.
Strangelove (1964) to The Shining (1980), of...
Daring Fireball
• John Gruber
Mac Mini and Mac Studio Supply Shortages
wsj.com/tech/personal-tech/apple-mac-mini-supply-3e7a7509?st=fKpr4Q
Nicole Nguyen, writing for The Wall Street Journal (gift link):
Mac Minis with larger-capacity RAM chips — a base M4 model with
32GB of RAM, starting at $999, and the M4 Pro models with 64GB of
RAM, starting at $1,999 — are “currently unavailable” on
Apple.com. And estim...
US and Tehran dispute the future of Iran's uranium
President Trump has again insisted that Tehran is ready to hand over its enriched uranium as part of a deal to end the war, despite Iran saying no such agreement is in place, and that its nuclear material is "going nowhere". Iran says the Strait of Hormuz is open - but shipp...
Adding a new content type to my blog-to-newsletter tool
Agentic Engineering Patterns >
Here's an example of a deceptively short prompt that got a lot of work done in a single shot.
First, some background. I send out a free Substack newsletter around once a week containing content copied-and-pasted from my blog. I'm effecti...
Daring Fireball
• John Gruber
Apple’s Developer Guidelines for Ratings and Review Prompts
developer.apple.com/design/human-interface-guidelines/ratings-and-reviews#Best-practices
Apple Design:
Avoid pestering people. Repeated rating requests can be
irritating, and may even negatively influence people’s opinion of
your app. Consider allowing at least a week or two between
requests, prompting again after people demonstrate additional
engagement wit...
Daring Fireball
• John Gruber
Follow-Up Regarding App Store Reviews, Which Are Definitely Busted
daringfireball.net/linked/2026/04/16/app-store-reviews-are-busted
I wrote yesterday:
And the apps that do the right thing — like Godier’s Current — and never solicit a review like a needy hustler are penalized.
On Mastodon, Steven Troughton-Smith responded:
Review prompts are the difference between a great app getting five
positi...
Europa Missions
Join us at PyCon US 2026 in Long Beach - we have new AI and security tracks this year
This year's PyCon US is coming up next month from May 13th to May 19th, with the core conference talks from Friday 15th to Sunday 17th and tutorial and sprint days either side. It's in Long Beach, California this year, the first time PyCon has come to the US West Coast since...
Charlie Puth on pop music complexity #piano #popmusic #shorts
Nancy Friedman notes the decline in quality in movie...
Nancy Friedman notes the decline in quality in movie taglines. “As movies have become louder, flashier, and more expensive, their taglines have atrophied: they’re limp, lackluster, and uninspiring.”
The 16th season of the Dissect podcast is a deep dive...
The 16th season of the Dissect podcast is a deep dive into Daft Punk; here’s the 1st episode.
Longreads
• Carolyn Wells
A Day with England’s Hunt Saboteurs
"Armed with drones, balaclavas and vegan sausage rolls."
Bodega Cats of New York
The Bodega Cats of New York project documents the working cats of NYC’s delis, bodegas, and corner stores.
The cat at the local deli wasn’t a pet. She knew the regulars. She kept the mice out. She gave people a reason to walk an extra block. And she was technically a violation of city health code.
That was six years ago. Since then, the project has documented over 150 shops, collected 13,500 petition signatures, and helped introduce the first legislation in New York City history to classify bodega cats as working animals.
Available soon in book form.
Tags: cats · NYC · photography