Some benefits can be reaped immediately (cyclomatic complexity), but others emerge over time (churn). A good example of this is say something like a 500 line file that has experienced 2500 lines of churn, and if that rate of churn is trending down or up. Surprisingly useful for understanding you've got a hotspot with an opportunity to pay down debt by spending more time on API design, or just breaking out whatever subset of the code is experiencing thrash.

The funny thing about complexity is that assuming you're lintting/formatting well, you can do a poor mans check by just looking at something like average indentation per line, deepest line, etc.

[1]: https://ouijit.com

As for Utopias/Dystopias, I don't know how to answer that without waxing poetic about ideas like chaos being the natural order, or beauty being in the eye of the beholder, .

Unfortunately I think this is inevitable because as a species we perform a steady march toward collective desires for items of science fiction, so start your adversarial double-think practice now.

Ive seen this as well, but at a certain point it just seems like if your code is online, someones either taking it wholesale or recreating core pieces from the spec regardless of any licensing.

Its a tough time to be building OSS if you don't have distribution solved AKA some pedigree from the before times.

On another note, the reason I was perusing the blog was because I was curious about more of how exe handles review/quality/testing because you mentioned there being no code review in another post about stripe billings (maybe a different author). So consider this a casual request for more content on reducing delivery bottlenecks :)

That aside aside, I agree. Since reading this article yesterday I've probably been overthinking an MIT from AGPL license switch for my own project Ouijit (shill time: https://ouijit.com). Its feels a little counterintuitive since AGPL encourages more open source downstream, but at the same time if I have solved some problem other agent harness devs are curious about, I just want them to take the solution without worrying about paying it back/forward.

codex(or whatever they're calling the bundled desktop experience) = selling inference

claude code = selling inference

conductor = selling hosted agents

superlogical = selling... something

I recommend everyone build their own. It's fun, you'll learn a bit, and assuming you work as a software engineer the dogfooding is built in. I built ouijit as my own example of this: https://ouijit.com & https://github.com/ouijit/ouijit (note that I am not trying to make money, just free as in libre AND beer)

Its intended as a terminal agent orchestrator aimed at power users who want a little more design than what I like to call 'hackerman' UXes.

- ask you for an email

- send you an email

- ask you for a username

- except you cant actually log in with this username directly

- im being forced to learn some new social url protocol

- why does the auth flow pass me through a new ui/url that seems owned by the project but visually disconnected (eg, different branding/colors for the form)

- my password manager couldnt bridge the gap

I'm notoriously fickle about dealing with signup/login friction, but the project sounds cool so hopefully my feedback is more actionable than curmudgeony.

[1] https://get.convictional.com/

The initial reason I built it was over dissatisfaction with how most agent orchestrators abstract stuff behind chat UIs, require login, or only support their companies own model/harness. I didn’t want to keep cobbling together scripts for worktree isolation between tasks though, so the yak shaving began. A couple weeks in, I received a message from a friend sharing that they just finished an 8 hour session working in it and they really liked it, and that they’d even shared it with their coworkers unprompted. I’ve been a software engineer/designer for ~10 years and its hard to put into words how rewarding that felt.

Some things I think are unique about Ouijit:

- Sandboxing is a first-class consideration, Ouijit supports running tasks in per-project Lima VMs. Right now I'm looking into making sandboxing pluggable so you can bring-your-own agent environment local or in the cloud, with a few baked in defaults.

- Integrated agents are automatically aware of a session-scoped CLI that enables them to orchestrate and drive tasks from any Ouijit terminal, with no additional setup.

- Tasks have lifecycle hooks that can trigger prompts, scripts, or both when tasks start, resume, move to review, or move to done.

- You can organize child terminals, markdown, and url previews to terminals. It's a little like Claude artifacts in that way.

Supports macOS and linux; integrates with Claude Code, Codex, Pi, and OpenCode agent harnesses.

Free and open source, no login, no telemetry.

Website: https://ouijit.com

Github : https://github.com/ouijit/ouijit

Comments URL: https://news.ycombinator.com/item?id=48732590

Points: 4

# Comments: 2

Prolix affected me at least once today, I appreciate this.

- I tried it on a repo that has very obvious boosting tells, like 99% of stars from week old or less accounts and it had higher trust signal than my own very much so real project. The issue to star ratio gave 5/5 despite the issues being core contributor created.

- In same vein as above, it seems to prefer open issues over closed ones in the scoring.

Have you considered scoring contributors/authors/issue creators separately to use as a proxy, instead of measuring the core things that are being gamed?