Agent tool design: less is more
Fewer, sharper tools beat a giant toolbox the model has to browse.
There is a version of this article that opens with theory. I decided against it, because the theory is easy to find and the motivation is not — so let's start with the thing that actually broke.
A short history
Tools matter less than the feedback loop they enable. A fast, boring loop beat every sophisticated setup I tried, and I've kept it ever since, even as the rest of the stack changed twice over.
The numbers below are from my own machine and my own workload. Treat them as a method, not a benchmark: the point is how the measurement was taken, not the exact figures it produced.
def load_config(path: str) -> dict:
with open(path, encoding="utf-8") as f:
return {k.strip(): v.strip() for k, v in
(line.split("=", 1) for line in f if "=" in line)}The design
What surprised me most was how small the final change was. Most of the work was understanding the problem precisely enough that the fix became obvious, and that understanding took far longer than the fix itself.
The first version worked, which is exactly why it was hard to see how wrong it was. Everything looked fine in the demo, and everything looked fine until traffic arrived on a Tuesday afternoon and the graph started doing something I'd never seen before.
Something the reader should not miss.
Implementation notes
I kept a log of every decision during this project, and reading it back a month later was humbling: the choices that aged well were the boring ones, and the clever ones all came with a footnote I hadn't written yet.
By the end of this post you should be able to reproduce the whole thing yourself in under an hour, including the parts I got wrong the first time — those are marked so you can skip them if you're in a hurry.
- Capture the current behavior before touching the code
- Make the smallest change that preserves it
- Delete the workaround you were sure you'd need
Failure modes
By the end of this post you should be able to reproduce the whole thing yourself in under an hour, including the parts I got wrong the first time — those are marked so you can skip them if you're in a hurry.
The first version worked, which is exactly why it was hard to see how wrong it was. Everything looked fine in the demo, and everything looked fine until traffic arrived on a Tuesday afternoon and the graph started doing something I'd never seen before.
Tools matter less than the feedback loop they enable. A fast, boring loop beat every sophisticated setup I tried, and I've kept it ever since, even as the rest of the stack changed twice over.