Surpasses. At least in my case, 27B tends to waste more time in the reasoning stage, too, so 27B overcomplicates. QFN just does a better job. Sometimes I eyeball, sometimes it's measurable like translation chrF++ 59.3 vs 67.5.
I switched from Claude code to pi about two months ago and can’t imagine having to go back.
However, this is very much a “build what you need” type of harness. You have to be willing to use other people’s extensions or build/vibe your own before the harness is useful.
Why I find pi better:
- multi-provider/inference API support out of the box
- greater visibility into token flow/usage
- minimal system prompt
- no hand holding/decisions made for you on what your agents can/cannot do
- extremely customizable
- stable, minimal bugs in the base harness that I’ve experienced
- minimal feature bloat
It’s a great harness for someone who wants extreme control and adaptability. I don’t recommend it to my peers and colleagues who aren’t into tinkering.
I have journeyed from Claude Code to OMP to Pi while working very hard to keep a super open mind, open to anything.
This from the post is why I landed on Pi
"What makes Pi the best fit here? Simple: it’s the only harness today that’s open source, vendor-agnostic, extensible, and minimal enough to not force opinions on you. It gives you full ownership and puts you in control. Being able to change the providers without switching apps alone is reason enough to adopt it, but the fact that it’s also so flexible makes it almost impossible for me to switch back."
I am building my own (intentionally simple) subagent orchestration framework inside Pi and am watching Pi Durable closely as I suspect it and Pi will unify before long and subagents are better supported in Durable. (Relatively uninformed opinion in this case, not using Durable yet.)
Haha. Locally it was passing. That badge is GitHub running the tests on its own machines, where a couple of new browser tests fail. Fix is going through the same loop: one agent fixes, another verifies, then it merges.
Yep, not looking to replace cyclomatic complexity. Looking to use it in what I've researched is a novel new way.
Instead of focusing solely on the complexity of the thing you change. Include the context around what you are changing also.
This allows the Change Impact formula to differentiate between two different files of the same length. One is a complex god class with high CC. The other is an entity module that is simple but just has a lot of methods due to properties. Adding to the god class really should warrant a refactor. Adding to the entity class is just another field to store for the entity. The Change Impact formula can differentiate and point to the god classes.
Yes. It’s called Severely Deficient Autobiographical Memory (SDAM).
I suffer from it quite badly. I can remember broad and important details, but struggle to remember even moderate details from anything beyond the past 2-3 months, with sharp falloff after that.
Once I started using Immich and was able to see my old photos from across adulthood and childhood (I used to just dump them onto HDDs), I experienced a sharp reconnection with myself and my past.
Maybe they could find something similar (Apple Photos, Google Photos, Immich, etc.)!
reply