Hacker Newsnew | past | comments | ask | show | jobs | submit | xmorse's commentslogin

OMP is a joke. don't use that garbage


Can you explain your opinion? I'm curious but such vague comments won't convince me.


Probably the same reason as oh-my-zsh, you don't need 90% of it. Further compounding the problem in an agent harness is that you are polluting the context window by throwing the kitchen sink at it.


I've been experimenting with omp because:

- it allows different models within one session via roles (I only have API, so pay per token)

- it's much more likely (ime) to use the LSP over grep for determining how code fits together

But I agree a 20k+ starting context is way overkill.

I find it's very hard to get information on harnesses people are using. I have to stay model agnostic so I avoid claude, codex, cursor, etc. I've used and tried opencode, which worked well, but obviously lacks the above features.

Does anyone have a resource for following what people are actually being productive with? With so much vibe going on it's hard to separate the wheat from the chaff.


Exactly! All these harness claims are just trust-me-bro, someone says omp is great for token saving, some other guy says it burns tokens...like dude which is it


this summarizes the average OMP user and dev

https://x.com/greg_horvay/status/2100764473392820433?s=20


Again, this is vagueposting, what do you have against snapcompaction? It seems to be a nifty way to save token costs. At least it's an interesting innovation. https://x.com/_can1357/article/2064802476742574459?lang=en


What a crazy thread lol. I really can't tell who is serious and who isn't there.


what do you use and why do you prefer it over omp


Just pi. `pi install` the packages you actually need or ask LLM to write a package for you. Keeping the harness minimal is the point of pi.


you are a joke if you think omp is a joke.


it's definitely not bigger. smaller if anything looking at how much faster it is


Finally! I have been waiting for this for years. Mojo has great potential. I would use it for everything if it was for me


Cerebras is very fast but you can basically never use it because of its scarcity


I am working on a project you can self host on Cloudflare with one command, to store secrets and passwords there. It has a cli similar to doppler

https://github.com/remorses/sigillo


usecomputer is a desktop automation CLI for AI agents. It works on macOS and Linux.

Screenshot, mouse control (move, click, drag, scroll), and keyboard synthesis (type and press) are all available as CLI commands backed by a native Zig binary


Does anyone really use these hacked up with duct tape skills? why not use something more reliable like playwriter.dev?


A compact, YAML-based inventory of your codebase, intended to be prepended to a coding agent's context at session start.

Purpose

- Give the agent a fast, structured overview of files and responsibilities

- Provide jump targets via top-level defs (functions/classes with line numbers)


Check out Critique too if you are looking for a side by side diffs TUI

It uses opentui, the same framework uses by opencode.

It can also render diffs to images, pdf and html. Very useful for agents to share diffs in remote environments like Openclaw or Kimaki

https://github.com/remorses/critique


Mainly using playwriter.dev to help debug CSS issues on the page. It's an extension you can enable in Chrome and let agents control the browser via CDP

https://github.com/remorses/playwriter


Interesting, thanks To your opinion, what's the benefits compared to the native Chrome remote debugging feature + the chrome-devtools MCP?


This one works as an extension so you don't need a new browser specifically for agents. It's easier to collaborate. Also if you run the MCP in non headless mode it brings to front the browser on every interaction, like opening a new page. With the extension this does not happen.

Another benefit is context usage.

The cli can also do a lot more than other MCPs because it uses code snippets run in a stateful sandbox to control the browser, so it can do virtually anything instead of just a few exposed tools like `scroll`, `click`


MCP eats lots of context (~20k tokens for chrome's). The more tokens you use needlessly, the faster your context rots (i.e., worse performance).


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: