Diego Petrucci

@diegopetrucci
50 Followers
260 Following
195 Posts
I mostly dance bachata.
To pay for it, I craft iOS apps.
Bloghttps://diegopetrucci.github.io/dev/
GitHubhttps://github.com/diegopetrucci
Xhttps://x.com/diegopetrucci

I've decided to try to reimplement the core components of OpenClaw in Swift & Vapor.

Follow along! https://diegopetrucci.github.io/dev/implementing-openclaw-in-swift-and-vapor-part-1-telegram-bot-and-llm-api

Implementing OpenClaw in Swift and Vapor, part 1: telegram bot & Anthropic APIs

A lot (!) of people are talking about OpenClaw, but not many seem to know how it actually works under the hood.

diegopetrucci/dev

@alex hey Alex, think you could find this useful for your newsletter? https://diegopetrucci.github.io/dev/self-improving-agents

I feel awkward being “spammy”, so this will be the first and last post I ping you for! Have recently started taking the dev blog seriously so I’m trying to get it seen by people (and criticism is super welcome)

A simple way to add memory to agents and make them self-improve

LLM agents are great, but them not having memory and not “learning on the job” is less than ideal. Take AGENTS.md, a hack borne out this need.

diegopetrucci/dev

Let's try something new this year: instead of just sharing my AI setup in private chats, I'll actually write about them and post them here. 🐗

The first one is documenting my super-simple way to have agents improve themselves automatically: https://diegopetrucci.github.io/dev/self-improving-agents

A simple way to add memory to agents and make them self-improve

LLM agents are great, but them not having memory and not “learning on the job” is less than ideal. Take AGENTS.md, a hack borne out this need.

diegopetrucci/dev

@fatbobman the examples are for a swift project, if that helps :D

and if you consider this spam let me know, i won't bother you again!

@fatbobman I think you might find this useful for your newsletter, even though it’s generic for AIs and not swift specific: https://diegopetrucci.github.io/dev/self-improving-agents

it's a way to have your agent self-improve by having it take notes automatically while you talk to it

A simple way to add memory to agents and make them self-improve

LLM agents are great, but them not having memory and not “learning on the job” is less than ideal. Take AGENTS.md, a hack borne out this need.

diegopetrucci/dev
GoodLinks 3 is here! Article and image sync across devices, Liquid Glass design with a new app icon, improved tagging with spaces and recent sorting, a new Quick Search palette, and more. https://apps.apple.com/app/id1474335294
GoodLinks App - App Store

Download GoodLinks by Ngoc Luu on the App Store. See screenshots, ratings and reviews, user tips, and more apps like GoodLinks.

App Store
@siracusa FWIW I’m still having issues purchasing in the TF builds for HS. Says that App Store region is wrong as you mentioned a few episodes back — tried the restore “trick” too and nothing happened either.

@brentsimmons Just to give you context, I use a popsocket to manage my RSI, and that kind of prevents me from being able to reach the top left corner easily. I’m not bothered for most apps that I use sporadically but I tend to have longer sessions with NNW.

This is not to say you have to change it, it’s a just FYI, as others might have the same issue :)

@brentsimmons just hopped on the beta, and it’s looking good! My only small piece of feedback though is that presenting modally the website breaks my “finger” flow — ie I can’t swipe back. It make using NNW single handedly very hard on bigger phones

Just donated.

This is a good cause, if you can, do the same! https://relayfm.social/@relay/113181730496772195

Relay (@[email protected])

Attached: 1 image Closing in on $780k raised for St. Jude with a lot of September left! https://stjude.org/relay

Mastodon