dlants.bsky.social
@dlants.bsky.social
Principal Engineer at Amplify, previously Desmos. Opinions are my own. he/him. I helped build teacher.desmos.com. blog: https://dlants.me masto: https://hachyderm.io/@dlants
The big win for me was replacing the puppeteer mcp server with the puppeteer skill. It exposes a programmatic interface (which means fewer round trips between the agent and the tool), and takes up a lot fewer tokens in my system prompt due to progressive disclosure!

github.com/dlants/dotfi...
dotfiles/magenta-skills/browser at main · dlants/dotfiles
my dotfiles for vim, zsh, i3, etc. Contribute to dlants/dotfiles development by creating an account on GitHub.
github.com
November 26, 2025 at 5:32 AM
I made the transition to using just ghostty after watching this video www.youtube.com/watch?v=o-qt...

I was able to replace my "project switcher" with hammerspoon github.com/dlants/dotfi... Still really miss some parts of tmux though, like github.com/ghostty-org/... though.
If Mitchell Hashimoto succeeds, Tmux & Zellij will disappear
YouTube video by Changelog
www.youtube.com
November 25, 2025 at 12:01 AM
After reading about it for a bit, it seems like Claude Code sprinkles in system reminders periodically to encourage the agent to use the skills jannesklaas.github.io/ai/2025/07/2...
Agent design lessons from Claude Code
First, acknowledge the obvious: The last post on this blog is from 2019. Most of the other posts here are about Uni. Some are still pretty good! Others maybe less so.
jannesklaas.github.io
November 15, 2025 at 7:58 PM
I imagine the real trick behind skills is whatever they're doing to get the agent to make use of the progressive disclosure efficiently. Maybe tuning? Maybe they periodically remind the agent that skills are available?
November 15, 2025 at 5:11 PM
Even prior to skills, I experimented with introducing progressive disclosure to my context md files - having it be a "directory", pointing to other md files that had more details about specific topics, but I find that the agent just ignores those links.
November 15, 2025 at 5:11 PM
Rational Funk with Dave King. youtu.be/XOTZttTVEqY?...

And that's how you get the gig.
RATIONAL FUNK Ep. 5: Individuality / Creating Tension / Brushes
YouTube video by RATIONAL FUNK with Dave King
youtu.be
November 8, 2025 at 2:59 PM
strongerbyscience is great for no-nonsense, evidence-backed nutrition advice. Some of it angles towards "getting ripped" but it is generally grounded in science and approaches it with a great 80/20 mentality. You might want to start from this guide: www.strongerbyscience.com/diet/
The Definitive Diet Setup Guide: How to Build and Adjust a Smart Nutrition Plan
This guide covers everything you need to know to set up a well-structured diet that supports your performance and body composition goals.
www.strongerbyscience.com
August 31, 2025 at 3:47 AM
They don't :(
August 21, 2025 at 2:16 AM
So the best way to get access to gpt-5 for this purpose is to set up a script that burns through $50 of tokens to get you to tier 2. That's absolutely bonkers! Just let me preload my account with $50 of credit to get access to the higher tier!
August 13, 2025 at 5:22 PM
30K TPM in tier 1 basically prevents you from being able to use a coding assistant. And there's no way to pay your way to a higher tier, you have to get there by organically spending $50
August 13, 2025 at 5:22 PM
I develop a coding assistant and OpenAI feels really hostile towards developers. Their sdk is a mess (does not adhere to jsonschema, is poorly documented).

Recently I tried to run some experiments on gpt-5 and immediately ran into their rate limits / tiering system.
August 13, 2025 at 5:22 PM
Open source tools should just let you define a regex for finding context files. (Like `autoContext` in magenta github.com/dlants/magen...)
August 13, 2025 at 5:14 PM
- I added authentication via anthropic's max and pro plans, which lets you use tokens the same way as claude code (thanks to the opencode folks for their open-source implementation for how to do this). This is much cheaper relative to using the pay-per-token API for heavier users.
August 13, 2025 at 5:09 PM
- I renamed "@compact" to "@fork", and also changed it from a forced tool use to just a regular tool. This means the agent can think before using it, improving the quality of the summary, and also this now takes advantage of the conversation cache, making these tool calls a lot faster and cheaper
August 13, 2025 at 5:09 PM
i tested it, it works
August 13, 2025 at 10:50 AM
Claude max auth by dlants · Pull Request #175 · dlants/magenta.nvim
github.com
August 13, 2025 at 10:03 AM
I can now have multiple ghostty windows open, one per tab, and use a command shortcut to quickly switch between them :tada:

Now I just need to figure out how to make my ghostty sessions persistent, and do some tweaking around how tabs are numbered, and I'll basically be there.
August 1, 2025 at 7:59 PM