I am not the author. But I think the idea is cool. I use this ruflo to handle it completely to avoid interrupt by AI when vibe coding.(not permission issue, I always use claude --dangerously-skip-permissions)

Is there anybody using it ?

It depends, case by case.

Companies like this don't deservce good enginners.

AnythingLLM ?

By the way, i checked your project "TermonMac", it's interesting, why not build one for Linux, right?

No, this is similar to "AnythingLLM", and i'll tell you why, this could work.

AnythingLLM requires about 2GB of RAM, just to be idle, and in the RAM Crisis, that this world is in today. i want to bring better optimization.

And, AnythingLLM utilizes a "UI-first, Logic-second" approach. Meanwhile, to combat this, i'm using a "Logic-first, UI-second" approach.

Not to mention, A system like AnythingLLM isn't easy to work on lower-end computers, and I for one, do not like that, because, i'm using a low-end computer to work on this.

The website is good but not good too.

Not easy to trust the choosed skill is work: - No input example - No output example

It's hard to play and trust the "skill".

I'm a software engineer building an app as a side project. I quickly realized AI bullshits a lot.

But you know, engineers bullshit each other all the time too. The difference is we have a way to verify it - logical chains. You have to build an argument that holds up before anayone buys in it.

So I though, can I make AI build its own logical chain ? Let it pass its own logic check before telling me the result.

That's how I created my own "think" skill. It's based on Meta's CoT paper: https://arxiv.org/abs/2501.04682

It roughly works like this: 1. FRAME - Challenge the question itself, hidden assumptions.

2. GROUND - Map what you know, what you need, what's missing.

3. ASSOCIATE - Launch multiple independent agents in parallel to generate hypotheses, avoid anchoring bias.

4. VERIFY - Break each hypothesis into atomic claims, verify each independently

5. CHAIN - Build a logical chain from survivors

6. PROVE and LOOP - Walk backwards from conclusion to premises, seearch for evidence, repair if broken

7. DELIVER - Start with "I was wrong if ...."

It helps me a lot. Whenever I need to check if Claude Opus 4.6 is bullshitting me. I just say "/think verify the above reasoning is correct" or "/think verify the above fix is correct and complete."

With vibe coding you don't really type code anymore. You just tell AI what to do in plain words(voice input helps a lot on phone) and it writes the code for you.

Would that make you reconsider coding remotely from your phone ?

I don't consider vibe coding to be coding :)

Unless, of course, you're reviewing and editing the code, in which case..... back to the problem of the phone.

What do you use for that? Any specific iOS/Android app ?

AI lies. The trick is making it double-check itself before it ships. I made a skill that forces AI to think more. Burns more tokens but improve accuracy. I suggest you use Meta CoT(Chain-of-Thought) https://arxiv.org/abs/2501.04682 to build a think skill to avoid this problem.

Thanks for the paper link! I'll check it out.

Good point. I agree with that. The key point is that heavy users benefit from this model while light users are basically subsidizing them. But it's a distribution when everyone shifts toward heavy usage, prices inevitably go up. The $17/mo pro price is already set to compete with other providers. Raising it would lose customers. Other tiers are also carefully priced to match competitors. So the only move left is to prevent the whole distribution from drifting toward heavier usage. That's exatly what this ban does.