Hey Hugging Face community
I’ve been working on an open-source initiative called PROTOCORE Signals – a lightweight system for generating multilingual, structured prompts for use in AI models, cognitive systems, or human reflection.
Why PROTOCORE?
Most prompt systems focus on questions or task-specific instructions.
PROTOCORE instead builds knowledge signals — composed of 4 structured parts:
- Context
- Vision
- Recommendation
- Question
These signals are written in JSON and can be rendered into HTML pages per language. They’re optimized for:
- Multilingual prompt sets
- Knowledge base prototyping
- Human+AI co-reflection
- Crawling/zettelkasten systems
Built with:
- PowerShell (for automation)
- JSON (as the data layer)
- HTML (for live web publishing)
Live Demo
GitHub
Sample Signal:
{
"lang": "en",
"context": "Rural areas often lack access to mental health services.",
"vision": "A decentralized AI assistant offering offline cognitive support.",
"recommendation": "Explore mobile-based AI models tailored for low-bandwidth zones.",
"question": "What ethical boundaries must be considered in AI mental support?"
}
[https://www.linkedin.com/in/osama-gangor-06715ab7/](https://www.linkedin.com/in/osama-gangor-06715ab7/)