AI-Native Engineer | Systems over Hype | Hermes • OpenCode • Codex | Anime & Manga
- Tweets 1,109
- Following 159
- Followers 208
- Likes 1,516
Pinned Tweet
Shipped something I wanted for a long time and early access to @typesafeai Jev made it possible.
Introducing JevX-Kit 🥳
I kept reading posts that felt AI-written and had no easy way to check. So instead of asking LLMs or myself, I built a tool, with JEV.
JevX-Kit is a browser extension for X with two powers:
1. JevX Pulse: scores posts while you scroll past, with an AI-written percentage, and informational tags. 🤖
2. JevX Block: hides Ads and Paid partnerships. No key, no network calls, runs entirely in the page.(No Jev for this, but Name consistency😅)
Why I trust it???
The tag set was first tuned on 100+, then 500+, and then 2000+ real posts for precision.⚡️
Scores are AI judgments, not verdicts, and they can be inaccurate...don't come after me please😶🌫️
Zero data retention, no telemetry, and your key never leaves the extension. 🔏
Chrome and Firefox builds are in store review, and taking longer than anticipated. 🥲
So, the code is now public, linked below.
What tag would change how you read your timeline?
2
2
3
248
So...@opencode web console got a few updates 😎
1. We got a new AUDIT tab now...for tracking workspace config changes. 🧐
2. The usage overview now shows % USED metric instead of % LEFT, and some color updates too for the same...this took me by surprise momentarily, ngl...usage went from red to yellow 🤣
@adamdotdev @alan__rice @iamdavidhill @jayair
Is it possible, somehow, to let the user decide if they want to see used/remaining% on their console? maybe with a toggle? 😅
1
2
534
YES please😭 as a toggle if possible
I am tired of typing
- point1
- point2
--point2.1
---point2.1.1
Or
- point1
- point1.1
- point1.1.1
15
When I opened the quoted PR yesterday, I was unsure if it would be useful enough to be merged to @NousResearch Hermes repo 😅
Then @Teknium looked at it, made a few tweaks, merged...and now it is the default `hermes status` behavior, while old one needs `hermes status --full` 🥹
P.S. If you don't see it yet, try after running a `hermes update`
Opened another PR for @NousResearch today, and it's for a CLI feature this time 💻
CLI was the surface that I fell in love with when I started using Hermes 🤩
But whenever I would run `hermes status` it would be a long o/p of text(117 lines for me)...consisting of every API key, providers, auths, messaging, etc....while all I wanted was just to see the ones connected 🫠
So i added a `--short` flag to the status command, which only shows the connected/configured ones...and now the o/p is only 8 lines(for my setup)....~93% reduction, and I still get the information I need. 😌
Hope it looks useful enough to be merged 😅
@Teknium @witcheer
1
20
Opened another PR for @NousResearch today, and it's for a CLI feature this time 💻
CLI was the surface that I fell in love with when I started using Hermes 🤩
But whenever I would run `hermes status` it would be a long o/p of text(117 lines for me)...consisting of every API key, providers, auths, messaging, etc....while all I wanted was just to see the ones connected 🫠
So i added a `--short` flag to the status command, which only shows the connected/configured ones...and now the o/p is only 8 lines(for my setup)....~93% reduction, and I still get the information I need. 😌
Hope it looks useful enough to be merged 😅
@Teknium @witcheer
9
37
4,863
You can follow the PR status at github.com/NousResearch/herm…
1
1
184
I do see the improvements you're making to @opencode web console...and I appreciate it 🚀
@jayair @alan__rice
2
17
750
Google releases(announces) Gemini Argon 4...their SOTA, and better than existing SOTA models
BUT...limited access to "set of trusted testers" 🤷♂️
$2/$10 per Million I/O
BUT...$4/$20 after introductory offer ends
Is this the comeback everyone expected? 🤔
Introducing Gemini 4 Argon – our new frontier model.
It’s built for complex workflows across coding, enterprise knowledge work, and cybersecurity defense – rolling out today to a set of trusted testers through our Fairwind Program.
51
Anthropic:
We all need to pace the frontier👉🏻👈🏻
Also Anthropic:
$518B investment, while $42B net loss🤷🏻♂️
BREAKING: Anthropic is making a massive bet that AI will transform the global economy more profoundly than industrialization, electricity and the internet, per Reuters.
Anthropic reported a net loss of $42 billion in 2025, and plans to spend $518 billion on cloud, computing and infrastructure obligations in coming year, according to the prospectus.
39
LFG 🚀🚀🚀🚀🚀
If someone is still sleeping on the @opencode Go sub...wake up NOW
1
103
Introducing: My daily browser ✨
- A browser that lives on Mac, Windows, Linux, Android, and iOS
- You already know how to use it 😎
- You mostly have it or it's flavors installed already and are happy 😇
- Sign-in, or don't...it still works
297
