
Counting the days until I get to go snowboarding. Need snow ASAP!

Founder @TradersPostInc. I like to build stuff. Trading automation software is my thing. Open-Source software PHP @doctrineproject @symfony.

Counting the days until I get to go snowboarding. Need snow ASAP!


Really excited about some of the stuff we are launching soon @TradersPostInc - MCP server for AI agents - Public REST API (personal and oauth apps) - TradersPost CLI - Trade journaling and analytics - Live market quotes and position/order updates streamed directly from your broker in the UI

Been a while since I’ve written a blog post, but I figured it was worth documenting the end of 25 years writing code: jasonmccreary.me/articles/i-…

This seems to work pretty well. And rest.sh/ for generating CLI.

What should I use to automatically generate SDKs in different programming languages for an OpenAPI specification? Is this the best choice? openapi-generator.tech/

Codex is unusable for me lately. 2 prompts in 5 minutes and session usage is done.




One of my favorite use cases for AI is coding loading states that match the loaded state 100%

First time using Twilio. Can't believe how janky everything is. Broken links everywhere and internal 500 server errors all throughout the app experience.

I don’t want to have to think about what model to use or how much effort to choose. Don’t make me think.
What is effort really? When do you change it it and why not just use max effort for everything? I dove deep into this problem, looking into evals and doing my own tests and I was quite surprised by the results.


I wanted to check out Upsun Dispatch. Searched and clicked and this is the first thing you get 😱 I hate that governments are forcing this kind of crap.

Been refining how we build software with agents at TradersPost. Agents build, test, review and iterate, with humans always in the loop reviewing throughout the process and giving final approval.

A big part of making agents super useful has been getting humans out of the middle of the agent feedback loop. Before, humans were the message bus. Now GitHub is the message bus. Agents stay on the PR, respond to events, iterate with each other and escalate to humans when needed.