Irakasi (@_Irakasi_) header
Irakasi (@_Irakasi_) profile picture

Using @tldraw to create my dreamed visual thinking tool. Scratching my own itch. Building in public. https://teapawt.org/t.co/v1a01KhpOz

Bilbao
Joined December 2023
Photo by Irakasi (@_Irakasi_)Photo by Irakasi (@_Irakasi_)Photo by Irakasi (@_Irakasi_)Photo by Irakasi (@_Irakasi_)Photo by Irakasi (@_Irakasi_)Photo by Irakasi (@_Irakasi_)Photo by Irakasi (@_Irakasi_)Photo by Irakasi (@_Irakasi_)Photo by Irakasi (@_Irakasi_)Photo by Irakasi (@_Irakasi_)
📣 Open call to agent builders: Let's read agent skills from `.agents/skills`, so people don't have to manage separate folders per agent. Today we pulled the trigger for Codex to read `.agents/skills`. Goal is to deprecate `.codex/skills`. Pls like/tag/RT for momentum.
166
497
158
3,501
464,967
Irakasi retweeted
👋 If you’re new to Codex, here are 7 beginner tips: 1. Start with: GPT-5.2-Codex high That is high reasoning. It is enough. Don’t be tempted with xhigh unless working on something really tricky. It uses more tokens and will be slower to finish. 2. Sometimes more reasoning may not help. You may need to give your agents better docs that are up to date. I prefer to have my agents create Markdown docs from DocSet that are local, instead of web scraping. I use DocSetQuery to create docs from Dash DocSet bundles. github.com/PaulSolt/DocSetQu… 3. Read @steipete post to get started. Bookmark his blog and follow him. Read his post, it’s gold, and so are his other workflow posts. steipete.me/posts/2025/shipp… 4. Copy aspects from Peter’s agents .md file and make it your own. There’s thousands of hours of learnings in his open source projects. github.com/steipete/agent-sc… Use the scripts too, things like committer for atomic commits are super powerful when multiple agents work in one folder. 5. Just talk to codex. You don't need complex rules. You don't need to create huge Plan .md files. You can get really good results by just working on one aspect of a feature at a time, handing it off, and then letting Codex do it. If you get bored waiting start up another project while you wait. Ask it to do something and then go back to the original one. Most likely it will be done unless you're doing a huge refactor. 6. You can always ask your agent to copy something from another project. Peter does this all the time and has agents leveraging work they’ve already done for new projects. I ask my agents to create Makefiles to build and run my apps. For new projects I have them copy the structure. See my workflow video: How I use Codex GPT 5.2 with Xcode (My Complete Workflow) youtu.be/o4iKnSYlhBQ 7. Ask it to do things … and most likely you’re going to need YOLO (danger mode) to get anything done without constant nagging. Enjoy your next app!
Which one? - Codex 5.2 high - Codex 5.2 xhigh - Codex 5.2-codex high - Codex 5.2-codex xhigh @steipete @mitsuhiko @badlogicgames @thsottiaux
29
77
9
1,084
238,257
We built Claude Code into a browser & whiteboard! It's called Decode @decodetool It’s the fastest & best way to give UX feedback to Claude Code while coding locally & enables it to review & test changes itself See thread👇to learn how whiteboards makes it an even better browser
55
96
31
1,341
162,848
Wow! @MatthewBerman covered uithub in his video a few days ago thanks to a great comment from @mattshumer_. Awesome! Got 6000+ new people onto the platform🤯
5
6
2
62
26,799
Irakasi retweeted
on our discord (link in bio) there are incredible things
6
12
272
48,549