Kun Chen

Former L8 engineer at Meta, Microsoft and Atlassian, now writing and building solo on agentic engineering. Writes Kun's Field Notes and posts a lot about AI coding agents on X.

Follow Full feed Add to list

Kun Chen did not write this page. What is this?

It collects the places they publish and what they have said there, each linked to the source. They have no account here. Is this you? Claim it, correct it, or ask us to remove it from ppll.

Something wrong on this page? Tell us.

Where they publish

Newsletter Kun's Field Notes Field notes on agentic engineering and solo building. Has a feed.
Practical essays on building software with coding agents, and on leaving a big-tech job to build solo. Over 4,000 subscribers.

Open blog.kunchenguid.com
Recent
- Your AGENTS.md is a Neural Net 23 Aug 2026 And here’s how to properly train it with gradient descent
- Kun's Pi Agent Config 1 Aug 2026 Minimal, practical, and elegant
- Evaluating the Effectiveness of Programming Languages for Agents 10 Jun 2026 And how to pick the best one
Show 7 more

Everyone Should Have an OPINIONS.md 5 May 2026 A game changer that wasn't possible without agents
The Month After Leaving Big Tech 4 May 2026 What it feels like to quit your job and go solo
How I built a starry night in TUI 28 Apr 2026 Twinkling stars, a moon strip, with a hand-rolled ANSI diff renderer
Org-Bench: Let’s Simulate the Org Charts Meme with Agents and See Who Wins 22 Apr 2026 And prepare to have your mind blown.
Making a Polished TUI Demo Video Without a Video Editor 21 Apr 2026 Recording, mocking, and polishing a terminal demo with off-the-shelf tools
How I Built a Reproducible Mac Setup with Nix 5 Apr 2026 A pragmatic setup that lets me get any new Mac into working shape in seconds instead of spending a weekend reinstalling everything.
Zero to One — Handbook for Entrepreneurial Engineers 2 Apr 2026 Ever since I was a teenager, my coding adventure has always centered around one question - I have great ideas. How do I turn them into reality? This article aims to give some answers to this question.
Link verified 8 Sept 2026. Recent items update automatically from the channel.
x.com @kunchenguid Where most of their writing on agentic engineering actually happens.
Short posts and threads on coding agents, developer tooling and what building solo with AI looks like day to day. Their most active channel.

Open x.com
Recent
Show 15 more

time for something fun and useful! 😀 introducing "/kun" - a skill that's a near-realtime distillation of myself https://t.co/5PisaaYVLE it runs on demand and helps you apply my knowledge, workflows and tools to think and build like a seasoned professional developer 6 Sept 2026
we’re entering an interesting age where AI makes repeating something that already exists drastically easier than creating something truly novel i worry the side effects are - 1. innovators who don’t have distribution get punished because someone with distribution can copy the innovation in an hour and take all the win if this happens repeatedly, people may simply stop innovating because innovation is already extremely risky. now with the added risk of getting stolen, ROI is a lot higher to just copy something that works 2. products with no network effect will all race to the bottom because an… 6 Sept 2026
one of the awkward things with controlling your terminal over ssh on your phone is all the keybinds that work great on a physical keyboard but is a pain to do on mobile in SSHHIP i can now do keybinds like shift+tab by just dragging the dial with one finger 6 Sept 2026
i forced myself to use a few non-mainstream models today. sharing my experience with everyone in case you're wondering 1. muse spark 1.3 i was genuinely surprised by how good this model is. if you silently swapped opus 4.8 with this without telling me, it'll probably take me quite a while to figure it out, and it would likely be from the communication style rather than capability at its contributor pricing, the ROI is quite incredible. i find very little motivation to even try deepseek v4 flash when this exists i don't like the phrase they keep using though - "intelligence that's too cheap to… 6 Sept 2026
have you noticed something new with claude 5 and gpt 6? this is the first generation where they released the biggest model fable and astra ahead of the smaller ones this is very interesting and worth speculating a bit the most reasonable explanation i can think of is that only fable and astra are actually “trained” opus 5, sonnet 5, and the upcoming gpt 6 sol, terra, luna are all “taught” this is NOT a good sign. why? have you seen opus 5 and sonnet 5? they are bad opus 4.6, 4.8 etc were “trained”, because there was never a mythos 4 to teach them, so opus back then was good it seems a trainin… 5 Sept 2026
@grok what is a CNC machine and what would an equivalent of it look like in “software factories”? 5 Sept 2026
there is a difference between openai's quota reset and anthropic's reset when openai resets limits, the 7-day timer starts fresh as well when anthropic resets, all the quota window timers remain as-is which one do you like more, and why? 5 Sept 2026
all the new model releases reminded me that you should update your CLAUDE.md and AGENTS.md after switching to the new models my recommendation is to keep what you have first, use it with the new models for a while, then run backpass https://t.co/DkM9b4TcZ0 to self-improve based on real transcripts 5 Sept 2026
a few minutes after astra started rolling out, fable limits got a reset should have seen that coming and used fable like there's no tomorrow! 4 Sept 2026
i was on a date with muse spark 1.3 at dinner when astra showed up at the window i killed muse spark on the spot and invited astra in the dinner continued 4 Sept 2026
one thing the recent model releases have consistently told us is that public benchmarks are completely useless now they reflect how well the models are trained to solve benchmark-like problems, but are far from how we actually work, because just like RLVR, benchmarks are inherently evaluating machines with machines, without human in the loop asking the model to build a cool 3d game is also a silly way of understanding model performance unless you are a game dev who thinks these vibe coded demos have any economic value LLMs are “spiky”. they can be really good at something very difficult, yet… 4 Sept 2026
after heavily testing out gpt-6-astra for a grand total of zero hours, my early first day read is - where is it??!! 4 Sept 2026
why is everyone announcing gpt-6 astra release except @openai? 3 Sept 2026
what on earth could take down gpt, claude and grok all AT THE SAME TIME????!!!! 3 Sept 2026
one of the biggest product mistakes in recent history is that Claude Design wasn’t built as a skill you can just run in Claude Code instead, we now have two separate systems to manage, creating an artificial boundary and unnecessary inconsistency 3 Sept 2026
Link verified 8 Sept 2026. Recent items update automatically from the channel.
YouTube Kun Chen Screen-recorded walkthroughs of their agentic dev setup. Has a feed.
Videos on their agentic engineering workflow and dev environment, aimed at other developers trying to build with coding agents.

Open youtube.com
Recent
- L8 Principal Building a Full Stack App with Agentic Engineering 27 Jul 2026
- Dad and Son Vibe Code an RPG Game 13 Jul 2026
- L8 Principal's Agentic Dev Environment From Scratch 5 Jul 2026
Show 12 more

L8 Principal's Agentic Engineering Workflow 20 Jun 2026
How small teams with AI move fast 6 Jun 2026
The circular ai money game 5 Jun 2026
Why middle management is getting hit 5 Jun 2026
Interest rate and layoffs 5 Jun 2026
LLM tokens aren't productivity 4 Jun 2026
ex-Meta Principal Explains the AI Token Game 3 Jun 2026
AI Tech Layoffs Explained by ex-Meta Principal Engineer 21 May 2026
First Month After Quitting My Big Tech Job as L8 Principal 17 May 2026
Don't Get Laid Off by AI (From a Lead Principal Engineer) 22 Oct 2023
Do you understand diffusion of responsibility? 26 Feb 2023
Don't get trapped by diffusion of responsibility 26 Feb 2023
Link verified 8 Sept 2026. Recent items update automatically from the channel.
GitHub kunchenguid Their code, including /kun and other agent-tooling projects.
Seventy-six public repositories, mostly small tools for working with coding agents -- CLIs, dotfiles, agent configs.

Open github.com
Recent
Show 15 more

New repokun — Think and build like a principal engineer. Leverage Kun's experience, knowledge, tools, workflows, and skills to help you get shit done. 8 Sept 2026 Commitsdoc: Add 'explainer' section to ENTRY.md · doc: Fix minor grammatical error in introduction · doc: Update heading to reflect principal engineer role
New repofly-with-me — Fly with me - watch a bird fly across an infinite gorgeous world. 8 Sept 2026 Commitsci: require no-mistakes attestation on PRs · doc: Update README with new description and links · feat: initial commit
firstmate — Talk to one agent. Ship with a crew. 8 Sept 2026 Commitsfix(procevent): reap orphaned runners and prevent launch storms (#3904) · feat(tests): run live harness guards by default when available (#3889) · fix(bin): allow pooled spawns without a git origin (#3885)
homebrew-tap — This repository is a general Homebrew tap for Kun Chen's apps and developer tools. 8 Sept 2026
treehouse — Manage worktrees without managing worktrees. 7 Sept 2026 Commitsfeat(pool): seed ignored files from .worktreeinclude (#129)
lavish-axi — HTML is the new markdown. Lavish is the new editor for your HTML artifacts. 7 Sept 2026
axi — Design principles for agent ergonomics. Higher accuracy with lower token cost than both MCP and regular CLI. 6 Sept 2026 Commitsdocs(catalog): last-resort port of #171 (trello-axi) (#172)
pi-launcher — A signed macOS launcher for your Pi agent. 6 Sept 2026 Commitsfix: stop hardcoding the last launcher version in the Pi-updater gate (#16)
grok-ship — Turn your Grok Bot into a software factory. 6 Sept 2026 CommitsMark grok-ship as superseded by the Firstmate/Grok Bot template (#5) · Add Firstmate-driven OSS triage pack (#3)
chrome-devtools-axi — The most agent-ergonomic browser automation 5 Sept 2026 Commitsfix: inject selected pageId on page-scoped MCP calls (#120) · fix(skill): defer installed guidance to the current CLI (#112) · ci: migrate no-mistakes gate to shared action (#111)
gnhf — Before I go to bed, I tell my agents: good night, have fun 4 Sept 2026 Commitsci: migrate no-mistakes enforcement to shared action (#211)
gh-axi — GitHub CLI for agents — designed with AXI (Agent eXperience Interface). 3 Sept 2026 Commitsfeat(commands): support issue and PR attachment uploads (#139) · fix(release): correctly promote and demote releases (#132) · fix: keep the installable skill as a CLI-deferring stub (#118)
vision — Agent skill that mines your repo's history to draft a VISION.md, stress-tests it with hard hypotheticals, and iterates with you on an interactive review board. 2 Sept 2026 CommitsMake VISION.md the only vision-alignment surface (#2) · doc: agent memory boilerplate · fix(vision board): cosmetic back sheets and a completion CTA (#1)
eddies-wallet — A virtual allowance and financial learning app for families. 2 Sept 2026 Commitsfix(app-review): reliably surface review outcomes to the captain (#136) · fix: add App Store privacy manifest (#135) · fix: finish allowance setup and keep amount fields visible (#133)
New repogh-axi-adhoc-sandbox — Fixed sandbox for gh-axi adhoc e2e (no createRepository in the daily path) 2 Sept 2026
Link verified 8 Sept 2026. Recent items update automatically from the channel.

Beliefs

Korrents What they believe 1 beliefs — each backed by an exact quote.
Each is a — compiled by korrents.com, not by them: the one-line wordings are korrents', the quotes are theirs.
Recent
- Unit tests have become just more source code to maintain — only end-to-end tests can be trusted to guard real behavior.
  
  unit tests are just source code at this point e2e test is the only real kind of test we can trust to guard actual behavior
  
  @kunchenguid on X Said 7 Sept 2026
  
  Full record — everyone who holds this →
All 1 beliefs →

What is a korrent?

A korrent is a belief a person has stated in their own words: one sentence stating the claim, backed by a quote and a source, kept at korrents.com.

Under a name here, the quoted block is what they actually said. The korrent beneath it is the claim those words support, in korrents' wording — tap it to see the record, its source, and who else holds it.

Nobody here wrote their own korrents. They are compiled from public statements, and a person can change their mind, which is recorded too.

Feed

As its own page →

Everything, newest first — across every channel.

8 September

Kun ChenGitHub 8 Sept

+ 4 more

7 September

Kun Chen 7 Sept

+ 1 more

x.com

Kun ChenKorrents 7 Sept

unit tests are just source code at this point e2e test is the only real kind of test we can trust to guard actual behavior

x.com

korrents.com
Unit tests have become just more source code to maintain — only end-to-end tests can be trusted to guard real behavior.

6 September

Kun Chen 6 Sept

+ 4 more

5 September

Kun Chen 5 Sept

+ 2 more

x.com
x.com

4 September

Kun Chen 4 Sept

x.com
x.com
x.com

+ 2 more

3 September

Kun Chen 3 Sept

+ 1 more

x.com

2 September

Kun ChenGitHub 2 Sept

23 August

Kun ChenNewsletter 23 Aug

Your AGENTS.md is a Neural Net And here’s how to properly train it with gradient descent

1 August

Kun ChenNewsletter 1 Aug

Kun's Pi Agent Config Minimal, practical, and elegant

27 July

Kun ChenYouTube 27 Jul

L8 Principal Building a Full Stack App with Agentic Engineering

10 June

Kun ChenNewsletter 10 Jun

Evaluating the Effectiveness of Programming Languages for Agents And how to pick the best one

6 June

Kun ChenYouTube 6 Jun

How small teams with AI move fast

5 June

Kun ChenYouTube 5 Jun

4 June

Kun ChenYouTube 4 Jun

LLM tokens aren't productivity

5 May

Kun ChenNewsletter 5 May

Everyone Should Have an OPINIONS.md A game changer that wasn't possible without agents

4 May

Kun ChenNewsletter 4 May

The Month After Leaving Big Tech What it feels like to quit your job and go solo

28 April

Kun ChenNewsletter 28 Apr

How I built a starry night in TUI Twinkling stars, a moon strip, with a hand-rolled ANSI diff renderer

22 April

Kun ChenNewsletter 22 Apr

Org-Bench: Let’s Simulate the Org Charts Meme with Agents and See Who Wins And prepare to have your mind blown.

21 April

Kun ChenNewsletter 21 Apr

Making a Polished TUI Demo Video Without a Video Editor Recording, mocking, and polishing a terminal demo with off-the-shelf tools

5 April

Kun ChenNewsletter 5 Apr

How I Built a Reproducible Mac Setup with Nix A pragmatic setup that lets me get any new Mac into working shape in seconds instead of spending a weekend reinstalling everything.

2 April

Kun ChenNewsletter 2 Apr

Zero to One — Handbook for Entrepreneurial Engineers Ever since I was a teenager, my coding adventure has always centered around one question - I have great ideas. How do I turn them into reality? This article aims to give some answers to this question.

26 February 2023

Kun ChenYouTube 26 Feb 2023

Do you understand diffusion of responsibility?