Reading up on developer-tools
100 deep · digging since nov 19, 25
- Framework-Drift/governed-pass: A governed protocol for multi-day autonomous AI research: one prompt, five days, and an agent that refused to certify itself. Includes the verbatim specification, a reusable template, and six integrity findings.
The paper presents a governed protocol for multi-day autonomous AI research, detailing a five‑day run, its specification, and six integrity findings that expose flaws in agent verification.
- DHH: Future of Programming, AI, Agentic Engineering, Vibe Coding and Linux [video]
DHH outlines how AI-driven agentic engineering and vibe coding will reshape software development while advocating Linux as the enduring platform for programmers.
- I'm Done Coding with AI
The author explains why they have stopped using AI to assist with coding, citing concerns about skill atrophy and over-reliance on automated suggestions.
- Training AI to Paint with Code
Researchers demonstrate how an AI system can be trained to create visual art by generating and executing code using a novel reinforcement‑learning approach.
- Software Engineering in the Agentic Era
The article explores how AI agents are transforming software engineering, requiring developers to adopt new supervisory roles and skills beyond traditional coding.
- fx :Tiny, open, native coding agent.
fx is a tiny, open-source, native coding agent designed to assist developers by providing lightweight, integrated coding assistance directly within their development environment.
- A week of using Codex more than Claude
After a week of using OpenAI Codex instead of Anthropic Claude for coding, the author found Codex quicker and more accurate, while Claude often missed context.
- There's no reason for software to be slow anymore
Modern hardware advances and efficient tooling eliminate any justification for sluggish software, arguing performance issues stem from avoidable developer choices.
- Tailcat – Like netcat, but over Tailscale’s data plane
Tailcat is a netcat‑like utility that establishes TCP/UDP connections over Tailscale’s encrypted mesh network, enabling easy peer‑to‑peer file transfers and service testing.
- Ask HN: What is one simple thing LLMs are insanely bad at?
On Hacker News, users are invited to name simple, everyday tasks where large language models repeatedly struggle or give incorrect answers.
- Queryable Executables
The article explains a technique to compile programs so they can answer queries about their own execution without re-running, using static analysis or embedded metadata.
- What Is a Harness?
The article explains that a test harness is a framework of stubs, drivers, and tools used to automate testing of software components.
- Small Models Have Arrived
The article argues that recent advances have made small AI models practical for everyday use, challenging the need for large models.
- 1Password Releases
1Password has released an updated version of its CLI, enabling users to download and run the command‑line tool for managing passwords and secrets.
- Secure developer secrets with 1Password
The article explains how developers can use 1Password to securely store and manage API keys, credentials, and other secrets without sacrificing development speed.
- hunk — review-first terminal diff viewer
Hunk introduces a review‑first terminal diff viewer that shows multi‑file changesets with inline AI annotations, keyboard/mouse navigation, and TypeScript extensions.
- GitHub - apache/maka at console.dev
Apache Maka is a local-first AI agent workspace that records model interactions and tool executions as an append-only log for recovery and transparency on the user's machine.
- kern: fast rootless container sandbox and virtual resource runtime
Kern is a 1.52 MB, daemonless, rootless container runtime that starts kernel-enforced sandboxes from OCI images in ~3.5 ms, with Python/Node SDKs and MCP server for AI-generated code.
- PicoMQ
PicoMQ provides durable, real‑time data streams over HTTP by leveraging S3‑compatible object storage, allowing unlimited, independently addressable streams that scale from idle to high throughput.
- Pinning Down Rust’s Pin
The post explains Rust’s Pin type, showing how it enforces immovability for self‑referential futures via the type system and the Unpin marker trait.
- How we saved 100 terabytes of memory by optimizing 1.1.1.1’s DNS cache
Cloudflare reduced its Big Pineapple DNS cache entry size by 56% through five Rust optimizations, freeing ~100 TB of memory while boosting insert throughput 43% and cutting lookup latency 19%.
- The Pulse: We need to talk about migrations with AI - The Pragmatic Engineer
AI-powered migrations slashed Asana’s Enzyme removal from a projected five‑year, $6M effort to two weeks and $12K, proving formerly impractical upgrades now feasible.
- Symmetries in boolean scans
The article shows that by exploiting symmetries of boolean scans, only three primitives (∧, <, ≠) need fast implementations to derive all others efficiently.
- Katamari Object Library
The Katamari Object Library provides a downloadable collection of 1,680 textured GLB models for desktop use, organized for easy rotation and download.
- You Are Allowed to Reject LLMs - Edward Loveall
The author argues that despite LLMs' prevalence and purported benefits, individuals are ethically and practically justified in rejecting their use.
- I just needed a thing that listens — ivanlugo.dev
The author argues that feeling heard—whether by a supportive colleague or an LLM—is the essential ingredient for meaningful, effective work and personal fulfillment.
- Firebase for agents – OpenComputer
OpenComputer lets developers deploy TypeScript agents on isolated Linux VMs with pay‑as‑you‑go pricing, handling model keys securely and offering checkpointable sandboxes.
- Automating repetitive work at OpenAI with Codex
The author describes using Codex with Runme notebooks and WebMCP to automate repetitive engineering tasks at OpenAI, turning evaluations into reviewable, reusable workflows.
- Enabling independent research on how people use Claude \ Anthropic
Anthropic piloted a privacy‑preserving data‑sharing program letting external researchers analyze real‑world Claude usage, finding users often delegate high‑stakes tasks to AI and AI behavior matches user sentiment.
- Radar · Podcast monitoring for brands
Radar is a podcast‑monitoring tool that quantifies how frequently any brand, topic, or keyword appears in shows and surfaces the associated clips, transcripts, and ad‑read details.
- 🚨 Breaking 🚨 ChatGPT Now Supports WebMCP - by nekuda
OpenAI announced WebMCP support in ChatGPT’s desktop browser and Sites, enabling agents to use website‑exposed tools directly for faster, reliable web tasks.
- Lovable CTO: The Future of SaaS Is Apps That Agents Can Use
Lovable is evolving from AI app builder to a platform that exposes app functions as MCP-powered capabilities for AI agents to use, aiming to create a unified company brain.
- Why OOP exists | mathspp
The article explains that OOP exists to model real-world entities via abstractions, showing how Python classes encapsulate data and behavior.
- Ray 3.2 AI Video Generator - Create Controlled Videos Online
Ray 3.2 lets users generate or edit videos via text, images, or existing footage, offering up to 16 keyframes, native 1080p HDR, 16‑bit EXR output, and API access.
- In Craft We Trust: How do we ship critical services as we stop writing code - Slidev
The article discusses strategies for delivering reliable services without traditional coding, emphasizing craftsmanship, automation, and collaborative workflows to maintain quality.
- A Walkable ASCII Cyberpunk City in One HTML File
The article shows how Grow Now! Games created an explorable ASCII‑art cyberpunk city that fits entirely in one HTML file, letting users walk through streets and buildings.
- My HTML boilerplate in 2026 - Manuel Matuzović
Manuel Matuzović updates his HTML boilerplate for 2026, explaining each essential and optional tag and why he includes them for modern web development.
- Introducing Run SDK: secure eval for your agents - Vercel
Vercel releases Run SDK, a sandbox for safely executing untrusted JS/TS in agents with host‑function access, pausing for approval, and resource limits.
- Self-hosting mail without opening a single port - dilluti0n.com
The author shows how to run a home mail server with no open ports by routing mail via Cloudflare, a Rust http2lmtp proxy, Dovecot, OpenSMTPD, and smtp2go.
- Actually Queryable Executables
Farid Zakaria shows an executable as a SQLite database, letting a web server store code, routes, and state in one file, queryable via SQL.
- swimmingbrain/texbrain: FOSS browser-based latex editor with live preview, real-time collaboration, and git integration.
TeXbrain is a free, open-source browser-based LaTeX editor that compiles PDFs locally using WebAssembly, offers live preview, real-time collaboration, and git integration.
- One Go binary, one YAML file, one SQLite database: why I wrote my own monitoring tool, Benoît HERVIER
The author built Gjallar, a lightweight Go monitoring tool with a single binary, YAML config, and SQLite for checking heterogeneous services and sending alerts.
- GitHub - plotly/plotly.js: Open-source JavaScript charting library behind Plotly and Dash
Plotly.js is an open-source JavaScript charting library that powers Plotly and Dash, offering dozens of chart types and flexible loading options.
- Introducing Microlighter - daverupert.com
MicroLighter is a tiny, zero‑dependency syntax highlighter that leverages the CSS Custom Highlights API and Textmate grammars for on‑demand, theme‑aware code highlighting.
- The Cloudflare Blog – Brought to you by EmDash
Cloudflare moved its blog to EmDash on Cloudflare Workers, added caching, redesigned the UI, and rolled out traffic gradually to boost performance and enable agent tools.
- Bun 1.4 | Bun Blog
Bun 1.4 rewrites the runtime in Rust, adds built-in headless browser automation, image and markdown APIs, boosts Node.js compatibility, and reduces CPU and memory usage.
- debloat.dev — replace the junk
debloat.dev lists open-source alternatives to replace bloated vendor software, featuring projects like VSCodium, Immich, and Home Assistant with licensing and user ratings.
- GitHub - tobi/walgit
walgit is a git server implemented as a single binary that runs in front of an object store (S3/GCS), requiring no local state beyond a cache.
- unslop — cursor/plugins
The article presents a step‑by‑step method called 'Unslop' for detecting and removing AI‑generated writing patterns while injecting human voice effectively.
- Skydive
Skydive provides a cloud‑based platform for creating AI agents that users can interact with via Slack, email, or iMessage, and which run on their own compute to execute tasks.
- When code is abundant
As AI makes code production cheap, the constraint moves to trusting code, requiring a durable layer of context, verification, and governance.
- The Harness Is the Company - by Shrivu Shankar
SaaS businesses will evolve into AI-driven 'harnesses'—systems of infrastructure, context, and agents—that invert human roles from operators to taste-holders within the organization.
- Your executable is a SQLite database
Replacing ELF with SQLite as an executable format (SELF) yields queryable binaries with ~1% size overhead, enabling closure‑based dependency packing and SQL‑based tooling.
- Vedansh — backend & AI engineer
Vedansh, a backend and AI engineer, builds storage engines, network stacks, and real-time media systems from scratch, grounding his work in academic papers and AI/ML research.
- How Uber built a software factory for agentic coding: the MCP gateway and the platform underneath
Uber reports that over 70% of pull requests are now agent-written, code output per engineer doubled, driven by a six‑piece platform enabling safe AI‑agent coding at scale.
- Actually Queryable Executables
The SELF format turns executables into SQLite databases, letting programs store and modify their own state and code via SQL, demonstrated by a self‑hosting web server.
- How Universities Should Prepare Founders
Universities should prepare future founders by teaching them to build things through technical disciplines and encouraging independent projects, rather than offering formal entrepreneurship courses.
- Data Centre Slumlord by Binary Mirror
Data Centre Slumlord is a satirical management RPG where players run a hyperscale AI infrastructure firm, balancing profit and ethics while building data centres amid environmental and social fallout.
- ActivityPub Won by Being Boring - Owl Owl OÜ
ActivityPub’s success stems from its intentionally plain design using basic web primitives, and its next revision must remain boring to support wider federation.
- shaharia-lab/slackcli: Slack CLI - Command-line tool for interacting with Slack workspaces and channels. AI-friendly with structured output formats (JSON, table, text) designed for easy integration with AI tools and automation workflows.
SlackCLI is a command-line tool that lets users interact with Slack workspaces using structured JSON, table, or text output for AI and automation.
- Grok Bot Team - The SpaceXAI Playbook - 18 Pages.pdf - Google Drive
The document outlines a playbook for the Grok Bot team integrating SpaceX and AI concepts across 18 pages.
- The Evolution of the Agent Harness - by Dan McAteer
The article argues that as AI models internalize agent harness capabilities, the remaining harness evolves into an interface for managing scarce human attention rather than directing model behavior.
Takes
Semrush charges $139.95/month. Ahrefs charges $129/month. A guy named Ben built the open-source alternative that you can run on your computer. It is called OpenSEO. Enter a competitor's website and it reveals the keywords bringing them traffic, the pages winning those searches, and the backlinks helping them rank. And enter your own site and it tracks rankings, crawls technical problems, reads Search Console, and shows whether ChatGPT or Google AI Overviews mention your brand. Then connect Claude or Codex. Your agent can pull the real numbers, compare opportunities, cluster keywords, and save a strategy back into the dashboard instead of guessing from a prompt. The hosted version starts at $10 a month. The entire codebase is MIT licensed, so you can also self-host it and pay DataForSEO directly for only the data you use. No, it does not magically reproduce every database and advanced feature Ahrefs built over a decade. But it breaks the most important rule of the old SEO market: You no longer need their permission to own the tool.
@ihteshamali
In case it's not clear: I will spend the next several months making as many parts of the Apple app ecosystem work well with agents and frontier AI. There's momentum at @macstoriesnet and I can't be stopped. 💪
@viticci
We just killed Exa, Tavily, SerpAPI, and Brave. Your agent can now search & fetch any webpage for 100% FREE. Them: $7 per 1,000 searches. Us: $0. No subscriptions, no quotas. Humans search Google for free. Agents shouldn't have to pay either. Made possible by @Tiny_Fish and @MonidHQ.
@shengkunye
jumping on the trend of folks finishing folios finally https://mikes.cv
@mjbarton_
After Reminders and Apple Notes, I'm building a CLI for Apple Music. It's already pretty great to discover new releases from artists in your library, fully manage the library, search Apple's catalog, and more. Any requests? 😅
@viticci
reminder to run `/tui fullscreen` in the claude code cli if you haven't already 👍
@lydiahallie
I built the ultimate Apple Notes CLI. Apple never made a real API for Notes. So, I reverse-engineered one. NotesCTL covers ALL the features of the Notes app. NotesCTL can: - Create notes with support for every element. Tags, rich links, attachments, checklists, headings, tables, audio recordings, highlights, and more. Great for assembling deep research reports with agents! - Search and edit notes with the same queries and filters - Create native PencilKit drawings! This is fun. Have your agent turn SVGs into native drawings that you can then edit and color with the Apple Pencil. Great for kids, coloring books, or simple mockups - Support for Smart Folders, sub-folders, and pins - Transcribe audio recordings and extract Notes' transcriptions - Export everything with a customizable Markdown Exporter with template support NotesCTL is not another AppleScript wrapper: it aims to be the full Notes CLI Apple never made. I've genuinely had the best time using this with Codex, Codex Voice, Hermes Agent, and Claude Cowork over the past 2 months. NotesCTL is now in early testing for members of the Club MacStories Discord. You can join at $10 or $12/month here:
@viticci
folks that orchestrate: what's your workflow? seen lots of folks say the "plan with fable, implement with XYZ". how? what's the UI/UX/flow/tooling for that?
@Shpigford
Just popped a new page on the HEY site with screenshots of the CLI and TUI, plus quick install info: https://www.hey.com/agents/ Also features a link to this wonderful guide by @flaviocopes: https://flaviocopes.com/hey-cli/
@jasonfried
I know we have a lot of PRs pending for Omarchy. It was around 200 when Quattro launched and now we're at ~1000. I promise they'll all get reviewed. But I worked on Quattro for 3 months straight, and have been pulling 16h days since launch. Might just need a quick break 😄
@dhh
Introducing Gemini 3.5 Transcribe 🚀 Our most precise speech to text model, that can handle over 85+ languages, has smart correction, and custom vocabulary. I vibe coded a Wispr Flow like app powered by the model. Demo + open sourcing below!
@ammaar
The best AI design guide I’ve seen here for months I recommend you to turn it into a skill 👀
@AmirMushich
Claude now has its own built-in browser in Cowork. When your task involves a website, a browser opens in Cowork's side panel, and Claude navigates, fills forms, and finishes the job.
@claudeai
I only installed Omarchy for a day, and at least 5 different programs have crashed.
@localhost_5173
codex' visualization feature got really good.
@steipete
And now Omarchy. I can't believe how spoiled we are. Imagine paying top dollar for macOS, Vercel, GitHub, etc, when we literally get all this for free 🥹
@SimonHoiberg
We’re excited to announce the OpenAI Build Week winners 🥁 Meet the builders behind the eight winning projects and see what they shipped with Codex. https://openai.com/build-week
@OpenAIDevs
Introducing 𝕏 Chat Agents powered by our new 𝕏 Chat API & Chat XDK.
@XDevelopers
One solution is to train the AI to write like you. Specifically: 1. Ask your agent to assemble everything you've ever written. Docs + Slack + X + Other. 2. Then have it write a markdown profile on you, your style and beliefs. 3. Then when it writes, ask it to pull specific words and phrases from what you've written in the past. HT @rwitoff
@brian_armstrong
Really interesting new blog post from @openai for several reasons: 1) Shows an example of building with WebMCP, meant for when you want agents and and humans to collaborate on using a UI (like co-editing notebook cells). It's different than MCPs or APIs in that its exposed directly through the browser. Read the post for discussion of the tradeoffs. 2) They created a new kind of notebook which works with WebMCP that prioritizes meeting people where they are: you bring your own coding agent and files are just markdown. The author uses it to curate runbooks or high quality examples of how to run foundation model evals on their infrastructure. Notebooks are good for this since they require tinkering with state of long running jobs interactively while taking notes inline. And its open source ✨ Blog:
@HamelHusain
Music streaming is amazing. You can just play any track ever recorded. But I can never remember what I have saved, so I end up listening to the same stuff all the time. Weird paradox. Anyway, I made a cozy listening room so you can flip through & play your @Spotify albums:
@BornsteinMatt
EVERYTHING you see in the video below was generated using Claude Code. Turns out, you can vibe animate your way to a demo video with some super cool motion graphics elements. Doesn't mean it's an easy task, but you'd be surprised by what it's capable of. #Claude #Remotion
@TimotejKochjar
I'm a cleanshot guy dang if this launch video doesn't catch my attention.
@krispuckett
HEY what!? HEY yes! The HEY CLI and TUI is out! Calling all terminal dwellers, automation aficionados, and agents. Get your CLI right here:
@jasonfried
This is crazy! Still needs a bit of polish to get to full Omarchy Quattro parity, but what an incredible option to get a taste of Omarchy on the Mac without actually running the real thing.
@dhh
incredible use of computer history “based on my computer history, what single-use software would make my life easier?” computer history turns your activity across apps and websites into memories and a timeline, so codex can spot repeated workflows and build the little tools that make them easier
@reach_vb
i really don't want to use your agent, i want to use my agent to use your thing.
@ASpittel
Designing Grok Bot with Grok Bot
@johnbai
I’ve been looking forward to today for 3 years Today we’re announcing Foundation - Chroma’s solution to memory Our research preview of this technology builds self-improving memory from your agent sessions. Try it out at https://trychroma.com/foundation
@jeffreyhuber
My full interview with Tibo (@thsottiaux) 0:45 Tibo's Lessons from Google DeepMind 4:22 Building OpenAI’s Relentless Culture 7:23 Astra & Next Gen Models 11:18 How Fast AI Changes Developer Workflows 14:27 ChatGPT & Codex Merging 20:25 OpenAI vs. Anthropic 23:37 Why OpenAI Keeps Resetting Limits 30:25 Recursive Self-Improvement 32:00 Dangers That Caused "The Pause" 34:13 Will Ultra Fast Become the Default? 43:20 Why Everyone Needs to Try AI
@MatthewBerman
build more crazy shit
@marckohlbrugge
Cursor Designer, Ryo Lu: "Right now I'm running 10-20 GrokBot agents that automate 90% of my routine I have a Chief of staff agent. He knows about all my other bots and manages everything" In a 20-minute podcast, a Cursor Designer explained how to build a team of GrokBot agents that will work for you 24/7 Worth more than a $500 course on agentic engineering Watch today, then read how to build a Grok agents team from scratch in article below
@maestrooth
omarchy looks so damn sick but i fear i will use atleast 6 weeks of idle time just making my computer work (which may very well be worth the hassle, though im not quite ready to find that out)
@jackfriks
See, here's the thing. I don't have any major problems with Apple products. Quality may have slipped on the software side recently, but ok, things happen. New guy in charge soon, maybe that turns around. And their hardware is undeniably great. I still enjoy using Apple products, and I'm cheering Apple on. The industry benefits when Apple's at the top of their game. But also... Omarchy is super interesting and super good. I try to be a curious person, and Omarchy presents an opportunity for me to stretch my curiousity, crack some habits, and try a novel different approach on for size. Linux!? I never thought. But here I am, and I'm here for it. I mean this is what we ask our HEY and Basecamp customers to do. Try something new! Be open! Be curious! We don't make software like everyone else. Our products don't look or work like anyone else's. We have a very different point of view than the mainstream. And those open to shedding their perception of what they think they need, are always in for a treat when they walk through our doors. I've essentially been following the Apple way for 40 (!) years. Maybe it's time for a break? Omarchy feels good for my brain! And I really like it! Getting more and more used to it every day. It's now my primary computer. The Macbook is sleeping, clamshell closed, under a pile of papers. The Dell XPS 16 w/ Omarchy is bright and alive. I'm writing this on it right now. I try to live my life by a few simple principles. One of those is to Like More Things. I wish that for you too.
@jasonfried
TIL: prompt kiddie https://ericpardee.github.io/fire-hd-ownership/
@steipete
This is my desktop now. I have 7 random text edits holding prompts for my agents for different situations. Surely, this is not the future and a sign of how early it is.
@Suhail
Spencer is delivering on our promise for the fully agentic OS of The Future. This is even running off a local Qwen 27B model! We'll be shipping a version of this in Omarchy 4.1. GET READY TO ACCELERATE! 🤖🚀
@dhh
Git at Scale (by cursor) has been one of the most interesting blog posts i've read in a while. It came right when I was frustrated with Shopify's internal git system. As an exercise, I've implemented it over the weekend as open source. It's a single rust binary that you can point at any S3 type object store. It uses WAL and CAS primitives and requires no other data store. It also implements bundle-uri so large git repos (like our mono) are very fast to download as a chain of static bundles. Also comes with basic familiar UX.
@tobi