Getting a little closer... Feeling a little dangerousy' right around the corner.

You can have no coding knowledge or you could be the best programmer in the world. You will learn a lot from this video. 4R Audit Video Great work Ry. Truly appreciate you!

Tough exam, lot's of reparation to get it ... Proud of myself, wanted to share it with the group. If someone is studying for this, and needs help, I'm glad to give you advice. Also happy to help if you're stuck with implementation.. just reach out. @Ry Mac .. I did it!!

Strands Harness does not help "ATSW_OS" because the system's architecture explicitly rejects the use of traditional multi-agent orchestration frameworks. Instead of relying on an external framework to manage agents, "ATSW_OS" uses the Interpretable Context Methodology (ICM). This method completely replaces framework-level orchestration with a standard filesystem structure. - Filesystem as Orchestrator: In "ATSW_OS", there is no orchestration framework installed; the folder structure itself is the program. Numbered folders dictate the execution order, and plain markdown files carry the prompts and instructions for each stage. - Handoffs and Context: Rather than relying on a framework for memory and context passing, intermediate work is handed off between stages using output/ directories. The output of one folder simply becomes the input for the next. - Avoiding Engineering Overhead: For sequential workflows that require a human to review the output at each step, agentic frameworks introduce unnecessary engineering complexity. By using a filesystem, every intermediate output is a plain text file that a human can easily read, edit, and save without needing to edit code or redeploy a framework. Because the folders handle stage sequencing and local Python scripts handle the mechanical tasks, an external agent framework like Strands Harness is entirely unnecessary for the stack.

What’s up fam?! I came across a clean, open-source project recently that I wanted to share for anyone experimenting with AI video workflows: Open Higgsfield AI https://github.com/Autom8AI/Open-Higgsfield-AI It’s an open source frontend wrapper that mimics the Higgsfield AI UI layout, giving you a pretty smooth, streamlined interface to manage prompts and generations. The code is pretty transparent (built on vanilla JS, Tailwind, and Vite), so you can audit it or self host easily if you’d rather your tools. It doesn’t have the friction of heavy platform bloat and gives you a dedicated UI wrapper to plug into your preferred API backends. If you like tinkering with custom frontend setups or want a solid layout blueprint to build your own workflow tools around, this may be worth a peek. Let me know if you end up testing it out or modifying the stack for any of your builds. I’m just learning and sharing as I move along. Bonus: Higgsfield released an api with $15 credit to start if you want to stick with the source.