Agentic AI writings
I love getting to share counter intutitve ideas of building agentic AI with other people and I mostly stick to twitter when I write about it. Here’s a small collection of my writings but you can find a lot more if you follow me on twitter.
Traps in building agents
This thread is about some seductive ideas that sounds very smart but are actually a massive waste of time when building agentic AI. You can read through a large discussion that happened on twitter about this
In building AI agents @cline , we've identified three mind viruses Mind Viruses are seductive ideas that sound smart, but don’t work in practice.
— Ara (@arafatkatze) August 19, 2025
1. Multi-Agent Orchestration
2. RAG (Retrieval Augmented Generation)
3. More Instructions = Better Results
Let's explore why! pic.twitter.com/p2p93Dt9QM
Agentic search
Agentic search is a very difficult unsolved problem and has historically had all different kinds of strategies. The most common ones are Grep and Vector search but each have their pitfalls. Its still unclear to me what the actual answers are here
Cursor vs. Cognition have opposite takes on agent search:
— Ara (@arafatkatze) November 12, 2025
- Cursor: Custom embeddings trained on agent traces improve accuracy by 12.5%
- Cognition: Embeddings are counterproductive, so we trained models to use grep with 8x parallel tool calls.
Both have benchmarks.
Hmmm.. https://t.co/QOKy6kvSVA pic.twitter.com/OhSKOTFMLw
Reverse Engineering AI tooling
Tons of AI tools are very similar but not quite the same, you can easily untangle by using this reverse proxy technique to take out their prompts
How to reverse engineer any closed-source AI coding tool prompts:
— Ara (@arafatkatze) December 19, 2025
This works for Claude Code, Windsurf, Copilot and literally anything by intercepting their calls
Method 1: Charles Proxy (for tools with internal endpoints)
Tools like Vscode Extension route to their own backend… pic.twitter.com/PbCDj5kwBM
Race to the RL Pipeline
Tons of AI startups would benefit greatly if they could fit into the RL pipeline workflow and this post reflects on how to position startups in this space
If you're building agentic tooling right now, there's only one finish line: the RL pipeline. Everything else is a race toward obsolescence.
— Ara (@arafatkatze) October 16, 2025
Looking at the landscape. A wave of startups has emerged around the atomic components of agentic systems:
- Agentic search ( @trychroma ,…
Open source model labs
This post is about open source model labs like MiniMax are actually doing extraordinary work with smaller models which are just as good as the larger ones
Many SOTA models have over a trillion parameters, but this only only has 10 Billion AAAAAND its open source.
— Ara (@arafatkatze) December 23, 2025
Hmm..... https://t.co/5J47EabcSf
Doing things and telling people
Many of us are just lurking through the internet and not sharing them thinking our ideas are not worth of sharing and this post explains why sharing ideas is critical not just for yourself but for the world
Hey @swyx question for you,
— Ara (@arafatkatze) December 23, 2025
You always say "do things, tell people." But I find the doing interferes with the telling.
I get lost in the doing. I have to forcefully block time on my calendar just to force myself to tell people by writing. And even then it feels like I'm spilling…