rahat (@Rahatcodes) header
rahat (@Rahatcodes) profile picture

Muslim | Building @animebento

Dallas, TX
Joined November 2011
Built a little tool for "vibe evals". Sometimes you just want to quickly see how a few models handle your prompts before testing them out in your app. Vibe Eval lets you test your prompts blind against different models. Built on OpenRouter vibe-eval.rahatcodes.com
2
8
456
Be nice to the AI
2
11
633
Ayyy side projects less expensive nowww
3
7
566
Updated the portfolio a bit and found some content/demos/workshops I liked from over the years. Nice little memory lane trip
2
3
456
Made some brand guidelines and I think claudes getting better at the thumbnail generation
7
511
claude code mods are dope built a fun one where every subagent gets its own tiny pixel-art squishy. watch them work, redirect them, change their model/effort, collect them all teams can build custom flows with mods too. it's just typescript hooks into your agents' lifecycle
5
1
7
650
I met Mokipe in squishys 🥟 · 9/48 species I found him while recording the demo for Squishyyyyys github.com/Rahat-ch/squishys
2
302
After building out evals on one of my latest projects, I wanted to go deeper and learn from folks who actually work in AI and build in it. Here's v1 of the free resources I'm using for my evals deep dive. Every author works in AI or is a researcher 👇
3
16
1,052
Lessgoooo
A little bit of night time hacking and experimenting with @n8n_io to see if I could recreate my Hermes Job pipeline.
16
713
A little bit of night time hacking and experimenting with @n8n_io to see if I could recreate my Hermes Job pipeline.
1
1
1
20
1,435