Latest story• Jul 1, 2026
Navigating Claude Code: Subagents Done Right
Learn how Claude Code subagents isolate expensive operations into separate context windows, keeping your main session lean and your token budget intact.
SMRTR Programming
Programming stories
Get hand-picked daily summaries of the best, most informative programming articles from around the web. Browse recent summaries, then jump to the original source when you want the full article.
Latest stories
Safari’s new MCP server lets coding agents inspect and debug websites
Apple's new Safari MCP server lets coding agents inspect pages, read console logs, and capture screenshots directly inside the browser.
How to Build a Browser-Based PDF Margin Tool Using JavaScript
Build a browser-based PDF margin tool with JavaScript and PDF-lib — no server needed, documents stay private, full custom control.
OpenWiki: CLI that writes and maintains agent documentation for your codebase
OpenWiki is a CLI that auto-generates and maintains codebase documentation, with GitHub Actions support for daily automated doc update PRs.
One Misplaced Click from Disaster: The Terrible UX of Developer Tools
When 'Delete Instance' lives next to 'Create Resource', engineers pay a silent mental tax that adds up fast.
Making a ‘Stylophone’ Music Synthesizer with Python and MIDI
Program an ESP32 as a MIDI keyboard in Python and connect it to GarageBand or other pro audio software.
SMRTR Programming
Get the next batch of curated stories in your inbox.
This archive is built from SMRTR newsletter stories. Subscribe for hand-picked stories without the extra noise.
More recent stories
The 20 Software Engineering Laws
Software development laws—like Brooks's, Hofstadter's, and Gall's—aren't about code; they're about human behavior. Recognizing these timeless principles before a project fails, not after, is what matters most.
The New Senior Developer Job Description: Half Engineer, Half AI Systems Architect
The senior developer role has split in two — and the new half involves architecting AI-assisted workflows, not just using AI tools personally.
Companies Are Making Claude and Codex Talk Like Cavemen to Stop AI’s Soaring Costs
Developers at OpenAI, Nvidia, and GitHub are using a caveman-speech plugin to force AI tools to respond tersely and slash token costs.
Behind the scenes of an AI-Driven Web Scraping System
AI-generated CSS selectors fail 30–40% of the time — here's how structured validation loops, DOM agents, and targeted retry feedback make scraping production-ready.
Reading Anthropic's "When AI Builds Itself" Changed How I Think About AI and Software Engineering
Anthropic reports Claude writes 80% of their production code, but the developers' real value — judgment and system thinking — remains irreplaceable.
Driving the Agent Quality Flywheel from Your Coding Agent
Google's Agent Quality Flywheel skill automates eval cycles for AI agents, catching subtle failures that look fine but silently give users wrong answers.
Claude Sonnet 5 is generally available for GitHub Copilot
Claude Sonnet 5 is now available in GitHub Copilot, bringing strong CLI performance, low latency, and Zero Data Retention across all plan tiers.
Ray Tracer in SQL
A single ClickHouse SQL SELECT query renders a full path-traced 3D scene with chrome lettering, terrain, and shadows — no UDFs needed.
GitHub Copilot now an Integrated Agent in JetBrains IDEs
GitHub Copilot is now natively built into JetBrains IDEs, offering a zero-config agent experience with OAuth login and CLI slash commands.
TakoVM – open-source sandboxing for your agent's code
Tako VM bundles Docker isolation, job queues, retries, and execution history into one self-hosted platform for running AI agent code safely.
How to Turn Any Color Palette Into a CSS Theme That Survives the Build
Four CSS rules for turning any color palette into a maintainable, accessible theme using custom properties and role-based naming.
Can a Product Designer Become a Game UX/UI Designer?
Game UX/UI design is about crafting player experiences, not just solving problems — here's how it differs from traditional Product Design.
Loop Engineering: Do Frontend and Fullstack Devs Actually Need It?
Loop engineering means giving an AI agent a goal and stop condition, letting it iterate autonomously until the work is done.
Why npm dependencies are a bigger security risk than your code
A compromised npm package deep in your dependency tree can steal CI secrets before your app even starts — here's how to reduce that risk.
I Hate AI, so I Built a Compiler
Frustrated by $2,000 in wasted AI tokens, one developer built Ampersand — a deterministic, graph-based language designed to make AI-assisted coding actually reliable.
Text Summarization With OpenAI and Ruby on Rails
Build a full AI summarization pipeline in Rails with OpenAI, covering chunking, Sidekiq background jobs, prompt engineering, and cost optimization.
Stop Overengineering: Build the MVP First, Optimize Later
Stop overengineering and ship your MVP first — a working product with average code beats a perfect project that never launches.
A field guide to the modern front end for developers who hand-wrote HTML
A 20-year technical deep-dive reveals how a simple "Buy" button evolved from a few HTML lines to 250,000 dependencies—each tool solving one problem while creating the next. Now, in 2026, the industry is swinging back toward...