Latest story• Aug 26, 2026
Herdr plugin to annotate anything in the terminal
Herdr Annotate lets you comment on terminal text selections and export annotated output as Markdown for use with AI agents.
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
How to evaluate session replay software: a developer's guide
A practical breakdown of session replay tool architecture — covering data capture, privacy masking, error integration, and AI-readiness — to help engineers pick the right tool.
Visualize whether you run too many coding agents, or too few
AgentPlayback visualizes your Codex and Claude Code agent sessions as a local timeline, showing activity, wait times, token usage, and estimated costs.
What is the quality of software that AI writes?
AI coding agents write working code, but often produce bloated, hard-to-simplify output with poor abstractions and unnecessary complexity.
A Thought on LLM Pair Programming
LLM-assisted coding creates an ownership vacuum — and someone will be forced to pay that debt, whether they want to or not.
The Node.js Event Loop, Explained Simply (with Examples)
Node's event loop explained: phase order, microtask priority, setImmediate vs setTimeout, and why blocking the loop kills performance.
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
Mdpreview – Markdown in the terminal (images, diagrams, big headings)
mdpreview renders Markdown in your terminal with real pixel images, Mermaid diagrams, and zero npm dependencies — just Node 22 or Bun.
Build Your Own AI-Powered Slack Bot with the Laravel AI SDK
Build a Laravel-powered Slack bot with swappable AI providers and full control over conversation memory using RAG and pgvector.
When Should You Raise an Exception in Ruby?
Use exceptions when a method can't honor its contract — not for expected outcomes that belong in normal control flow.
The End of Programming
AI rewrote Bun's entire JavaScript runtime—one million lines of code in eleven days—for ~$165,000, with minimal human involvement. As AI-generated code surges exponentially, traditional human-written programming may already be...
Your Job Ends at 5. Your Developer Brain Doesn't.
Developers struggle to mentally clock out because unfinished code has no natural stopping point — a simple written shutdown ritual can help.
Maiao: Gerrit-style code review workflow for GitHub, GitLab, Gitea, others
Maiao adds a `git review` command that turns each commit into a stacked PR/MR, managing dependencies and fixups automatically across major git providers.
Iterize IDE – open-source multi agentic coding platform
A personal research hub showcasing original projects spanning symbolic AI cognition, new linguistic theory, an agentic IDE, and experimental AI reasoning architecture.
What Replacing React with Hotwire Really Costs
Replacing React + Redux with Rails Hotwire cuts bundle size dramatically, but requires rethinking every screen rather than translating code line-for-line.
Slash-tokens – know LLM cost before the call leaves your machine
Intercept LLM API calls before they happen and estimate token costs in sub-millisecond preflight checks using a tiny 4.8 KB WASM engine.
A Manifesto for Responsible Agentic Coding
A practical framework for responsible AI-assisted development: keep humans accountable, review LLM output often, and never sacrifice code quality for speed.
Beyond Vibe Coding: A Quick Field Guide to Agentic Engineering
Agentic Engineering replaces chaotic AI prompting with disciplined architecture, decomposed tasks, and TDD — turning AI into a reliable development tool.
LatticeDB – Like SQLite but for graph databases
LatticeDB packs graph traversal, vector search, and full-text search into a single embedded file with no server and zero config.
How I Use Claude Code
Use a pre-commit hook to re-inject your CLAUDE.md rules into Claude's context and catch code quality violations before they're committed.
The Move to Python 3 Begins
EVE Online is finally migrating 2.4 million lines of Python 2.7 code to Python 3, promising better performance and modern tooling.
What Do You Do While AI Codes?
When AI is busy coding for 5-20 minutes, smart developers use the gaps for learning, paperwork, or deeper problem thinking.
LIGH – Let coding agents use and verify the iOS apps they build
LIGH lets AI coding agents autonomously run and verify iOS apps on Simulator — 3.6× faster and 7× fewer LLM tokens than vision-based tools.
Your React app doesn’t need all that JavaScript
Six common React UI patterns replaced with native HTML and CSS equivalents — less JavaScript, fewer re-renders, and simpler code to maintain.
Inside the Mind of a QA Testing Agent
QA.tech's AI agents test by user goals using a knowledge graph, adapting when UIs change instead of breaking like selector-based scripts.