PWV Portfolio Feed

every day, we stray further, etc

Put differently: specs as code

This could look something like "literate TypeScript". A mixture of TypeScript, Markdown and React. Think Jupyter Notebooks but modern and embracing web technologies.

Still early days but I'm already working towards this spec-driven vision.

I'm slowly but surely forming a better idea how to do "better specs".

The key idea is to make them structured and allowing a spec to contain structured source of truth definitions and rich artifacts such as code-defined diagrams, API/CLI/schema definitions, etc.

Tomorrow Command Code will drop one of the best deals ever.
Noon 12pm pacific. Guess and it's yours for free.

Tomorrow Command Code drops one of the best deals ever.
Noon 12pm pacific. Can't wait to share.

This is getting ridiculous!

MiMo V2.5 is up to 99% off on Command Code.

Let's get it!

we're growing too, hiring in sf and remote worldwide.

check open roles on my profile bio on X.

if you like engineering deep dives on how we're doing all this, i've linked some relevant posts below.

try it now.

npm i -g command-code

we're going open source next month. today we're a cli at the core, and we're also launching a full-fledged gui app, sandboxed background agents, and cooking up something fun i can't wait to share.

we have a big roadmap ahead of us. the feedback we hear most is that Command Code feels fundamentally different: an approach built on taste and repair.

what's next.

we've applied the same repair idea to ai design slop, and bundled a /design capability² so every developer can level up their design work. the early response has been great.

i believe these numbers are a consequence of getting those two things right.

net effect: developers using Command are writing production-quality code on open models, 10x to 100x cheaper, without fighting tool calls, while building repo and team-wide coding taste that compounds.

Command Code builds your coding taste into skills, learned from your accepts, rejects, edits, prompts, and the corrections you repeat. over time it drifts away from generic code and toward how you actually ship code. it learns continuously, and while it is early, the direction feels right.

on the second idea: taste.

developers have either adopted Command Code or used the same idea to build repair harnesses for nearly every top coding agent. i take that as more meaningful validation than anything we could say about ourselves.

i wrote the idea up openly¹ a few weeks ago, and it has quietly become a de facto way people fix open models.

open-model tool-call failures are not deep, they are a small finite set of contract mismatches. so we repair them, with zero token loss. what started as 4 repairs is now the largest repair layer in the space: 36k tool-call fix variants.

on the first idea: open models.

we fixed the "open models aren't good enough at tool calling" problem. our research came down to two things, quality and speed, and both trace back to one root cause: broken tool-calls that open models produce, especially when you use a bad harness.

we're building for taste and developer experience. so instead of making a soup of thousands of models, we build for the best ones, open or closed. the goal: a coding agent that feels like an iphone, opinionated and with taste, not a random android or a windows phone with no taste.

Command Code is built around two ideas:

1. open models should be production-grade for coding. 

2. your coding agent should learn your taste.

BIG day for us!! Command Code has crossed $1M in annual run rate, 1 trillion tokens of usage, with over 9K customers, just 24 days after beta launch.

We believe this makes it the fastest-growing coding agent harness for open models. 3rd largest by usage.