Mandar Zope

Harmony, not the peace, is the true solution.

[Mandar Zope @mandarzope]

We need Human Control.
https://t.co/zLVU7q0Jxm

[Mandar Zope @mandarzope]

@steipete With getting frequent model capacity errors we have shifted to manual coding 😂

[Mandar Zope @mandarzope]

LLM war is tough. I tried cancelling the OpenAI subscription and got one month free.

In Cursor I got $10 off on current billing.

[Mandar Zope @mandarzope]

If Anthropic will never train on this data then why do they want it for human harm prevention process. I mean aren't they indirectly harming humans.

[Mandar Zope @mandarzope]

@TheSpacerr Generating images

[Mandar Zope @mandarzope]

https://t.co/0wWxFtBjD6

this repo seems gold mine

[Mandar Zope @mandarzope]

What's happening.

Cursor - At Capacity
GPT-5.6-sol - Selected model is at capacity. Please try a different model

[Mandar Zope @mandarzope]

@AbhiCodes15 Discord

[Mandar Zope @mandarzope]

@thsottiaux Been using OpenAI, and I like how creative these models are. Models like Fable or Claude Code seem designed for small prompts, but once you give them a complex, large problem, they start to miss details that Sol-Max captures easily in an existing large project

[Mandar Zope @mandarzope]

The real magic happens when you transition from running one-off prompts to building a closed-loop feedback system. Your workflow should look completely different today than it did eight weeks ago. If it doesn't, it's time to take the training wheels off.

[Mandar Zope @mandarzope]

Skills like 'superpowers' and 'grill-me' are just for beginners. After 30 days, if you aren't constantly updating your AI frameworks based on long-term memory, you're missing out on a massive wave of personalization and efficiency.

[Mandar Zope @mandarzope]

@TTrimoreau Claude

[Mandar Zope @mandarzope]

@_itsjustshubh Also it says "Say a word and I'll implement it" and later that burns 10% of your $20 plan. :)

[Mandar Zope @mandarzope]

The "Superpowers" plugin for Claude Code is just a clever way to turn a $0.05 bug fix into a $15 token bill.
Change my mind

[Mandar Zope @mandarzope]

No one is checking if the AI Agents are GDPR or CCPA compliant.

[Gergely Orosz @GergelyOrosz]

I got messages from concerned devs how their codebase was uploaded without their knowledge or consent via Grok CLI (from SpaceX).

It seems that SpaceX sneakily uploaded this code for lots of users and customers… absolutely unacceptable IMO

Trust burnt like there’s no tomorrow

[Mandar Zope @mandarzope]

@ClaudeDevs I just open-sourced Fullstack Swarm, a setup that distills agent loops into clear, reusable skills. It's highly beginner-friendly to get started with, and modifying the loops is completely frictionless.

https://t.co/Xx8oIMZ0c7

[Mandar Zope @mandarzope]

@AnatoliKopadze Love this thread! I have compiled my agent loops into skills and agents. I'd honestly love any feedback on the setup or architecture if you have a second to glance at it!
💻 https://t.co/Xx8oIMZ0c7

[Mandar Zope @mandarzope]

Nice @OpenAI has removed 5 hour limit. It was major blocker for using SOL-max or ultra

[Mandar Zope @mandarzope]

Has anyone tried it yet? How is the experience? 👇

https://t.co/h9MdjXnKRM

[Mandar Zope @mandarzope]

For planning a large feature on gpt-5.6-sol Ultra ate Fable 5.