DEV Community: my god

I finally build my first AI-native web operating system but with some work remained 😅 github.com/hjk-maker/Solips-webos

my god — Mon, 13 Jul 2026 10:14:36 +0000

I finally build my first AI-native web operating system but with some work remained 😅 https://github.com/hjk-maker/Solips-webos

my god — Mon, 13 Jul 2026 10:13:58 +0000

GitHub - hjk-maker/Solips-webos: Solips-webos is a browser-based web operating system with AI at its core, no backend or installation required. Includes Windows, File System, Terminal, Apps, and LLM from multiple providers (including Ollama, OpenAI, etc). Offline, Extensible, Static Hostable · GitHub

Solips-webos is a browser-based web operating system with AI at its core, no backend or installation required. Includes Windows, File System, Terminal, Apps, and LLM from multiple providers (includ...

github.com

Alright, here goes - I've been working on this project for months now and it's time that I get another pair of eyes on it. I created Solips-webos, a browser-based OS with an AI kernel. It's completely client-side with no backend to it. The idea itself wor

my god — Mon, 13 Jul 2026 10:08:16 +0000

my god

Jul 12

I have created something that I believe in, but I cannot create by myself, I need assistance to test my web operating system.

#ai #webdev #programming #javascript

1 min read

This project is dead if there’s no community involvement – can you spend 15 minutes testing it?

my god — Sun, 12 Jul 2026 11:35:34 +0000

Let me be real, I have created something cool and hit that thing
that kills most one-man shows – I need testers, but they are not there.

Solips-webos is a desktop OS running entirely in your browser,
controlling it with an artificial intelligence, no server needed,
pure static files.

What I need (and why I need to go public):

Bug reports - I tested it myself but everyone knows that you are blind in some places
Usability feedbacks - What is hard to understand, what is broken?
Do you think it is helpful or just some fancy technology show?

What I ask is little:

Clone the repo
Open index.html
Use the browser like a regular OS
Report issues when something will go wrong (and something will)

Why I am being this bold:
Because I was in your shoes – browsing through developer posts
asking for help. And I understand how everyone is looking for some help
but no one has time to do so.

However, this project is important to me. I am sure that the future
of browser-based OSes with AI integration is bright and I can not prove
it otherwise.
If you answer yes:

I'll personally get back to you on your feedback
I'll address any issues you spot
I'll give you credit in the README
You'll be influencing the development of this thing

Repo: https://github.com/hjk-maker/Solips-webos

It's nice to know that there are people out there who actually care enough to look at this.
But if you ARE able to help - please do.

How much can be done on the client side? Created a web operating system as a demonstration.

my god — Sun, 12 Jul 2026 11:28:30 +0000

Have been working on an architectural problem: how much of a traditional
OS can we move into the browser?

In my current project I created an experimental browser-based OS which
includes:

Window manager, file system, terminal
AI kernel with tool-calling abilities
No backend (only indexedDB/localStorage)

The experiment seems to be successful, but what really matters is the
discussion.

Questions are:

Architecture: Pure client-side approach to solve some problems or do we fight the platform?
AI integration: The OS has AI which is capable to control the UI via virtual cursor. Useful pattern or just overengineered?
State management: IndexedDB + localStorage for all - scalable approach or a road to hell?
Extensibility: Added an app registry system. Any patterns to share with developers of plugin architectures?

This is not an issue thread about bug-fixing. Core works well. It's more
like architecture discussion: is there any sense of going this way.

Project repo: https://github.com/hjk-maker/Solips-webos

So, what do you think about thick-client web apps?

I have created something that I believe in, but I cannot create by myself, I need assistance to test my web operating system.

my god — Sun, 12 Jul 2026 11:22:34 +0000

Alright, here goes - I've been working on this project for months now and
it's time that I get another pair of eyes on it.

I created Solips-webos, a browser-based OS with an AI kernel.
It's completely client-side with no backend to it. The idea itself works,
but:

I know there are bugs. I know it requires some testing. And I'm unable
to catch everything myself.

What I need:

10 mins of breaking my software by just clicking everywhere.
Help finding those edge cases I'm unable to catch myself.
Pointing out what is unclear and doesn't work.

I don't need any code review or advice on architecture (if anyone wishes).
I need someone to use it and to tell me how it's failing.

Repo: https://github.com/hjk-maker/Solips-webos

All you have to do is just to open index.html and start breaking it.
That's all.

Anyone who was in my position, while building a project -
knows exactly what I mean.

I'll personally reply to every issue raised. I'll solve those problems.
I just need you to look at it with fresh eyes.

Can you help me improve it?

Created a web OS that runs 100% in the browser with an AI kernel, no backend, zero install

my god — Sun, 12 Jul 2026 11:10:48 +0000

Hi folks,

I have been trying to explore the edges of what is possible with
browser technology. I am presenting my latest creation: Solips-webos - 100%
client-side desktop OS.

The idea: Imagine a Large Language Model actually using an operating
system instead of just chatting about it. Clicking windows, managing files,
executing commands, within a secure browser environment.

Technological aspects:

Pure static HTML/CSS/JS (just open index.html)
Persistent data storage using IndexedDB
AI kernel with tool-calling capabilities (Ollama, OpenAI, Anthropic, Gemini, Grok)
AI-controlled virtual cursor
Sharing of an LLM between browser tabs using BroadcastChannel
App registry and file association system

The cool part: The AI controls the OS using a visible cursor - not
only calling APIs, but manipulating the UI like an actual user would.

No server, no build process, no install. Curious to see what everyone
thinks about this way of building browser OS.

GitHub: https://github.com/hjk-maker/Solips-webos

What would you make of this system?

Notion

my god — Sun, 15 Mar 2026 03:49:07 +0000

HJKSwitch — Build Your Own Virtual Wi-Fi Network with AI, Encryption & Futuristic GUI

my god — Sat, 14 Mar 2026 19:31:58 +0000

HJKSwitch – Virtual Wi-Fi, AI Traffic Intelligence & Futuristic Networking in One Python File

Imagine turning almost any computer (Linux or Windows) into a:

Virtual Wi-Fi access point / router / switch / mesh node
AI-powered traffic classifier & anomaly detector
Encrypted P2P file & message sharing hub
Parental control + firewall + IDS device
All with a beautiful custom Tkinter dashboard, web control panel, and even PDF export of the source code

…and doing all of this in one single Python file.

That’s exactly what HJKSwitch is.

GitHub →https://github.com/HJKEnsignElectronics/HJKEnsignElectronics

(Replace yourusername with your actual GitHub username)

Why did I build this?

I wanted to experiment with many networking, security and AI ideas in one place without creating 15 different repositories.

So I ended up with a monster single-file project that currently implements 100 planned features (yes, really — see the roadmap inside the code).

Some of the highlights:

Virtual Wi-Fi AP (hostapd / netsh hostednetwork)
Dynamic DHCP server
NAT / Internet sharing
AES packet encryption + optional custom encryption modules
Per-device traffic shaping (tc on Linux)
AI traffic classification (small PyTorch model)
Intrusion detection (SYN/ICMP flood, high packet rate)
Parental control (domain whitelist)
Tkinter GUI dashboard with network map (matplotlib)
Web control panel (simple HTTP server)
CLI & voice-command-like text interface
PDF export of the entire source code (!)
Particles.js futuristic website (GitHub Pages)

Quick start (host mode)


bash
# 1. Generate self-signed cert (needed for TLS file/messaging)
openssl req -x509 -newkey rsa:2048 -keyout hjkswitch_key.pem -out hjkswitch_cert.pem -days 365 -nodes -subj '/CN=HJKSwitch'

# 2. Run as host (needs admin/sudo)
python hjkswitch_combined.py --mode host
Client mode is also supported:
Bashpython hjkswitch_combined.py --mode client
Current status (March 2025)

what is this? hmm? it is !

my god — Thu, 19 Feb 2026 18:25:56 +0000

HJK-Quantum: A Unified Quantum- Computational Framework for Adaptive Intelligence
Author: Hjk Maker
Affiliation: Independent Researcher
Date: 2026
github:https://hjk-inc.github.io/HJK-QUANTUM-
Abstract
HJK-Quantum is a conceptual research framework that unifies quantum-inspired computation, adaptive intelligence, and mathematical programming into a single scalable architecture. The model introduces quantum-state-driven information encoding, probabilistic reasoning layers, and self- optimizing learning dynamics. Unlike conventional quantum computing, HJK-Quantum does not require physical qubits; instead, it employs quantum-formal abstractions
implemented on classical or hybrid hardware. This paper is prepared to meet student research submission standards, emphasizing originality, clarity, and ethical compliance suitable for academic review platforms such as Achieve.org.
Keywords: HJK-Quantum, quantum-inspired computing, adaptive intelligence, probabilistic systems, student research

Introduction Modern computation faces fundamental limits when addressing uncertainty, combinatorial explosion, and adaptive reasoning. Quantum computing promises solutions but remains constrained by hardware limitations. HJK- Quantum proposes an alternative: a quantum-inspired computational intelligence system that integrates principles of superposition, entanglement, and measurement into software-defined architectures. The goal of HJK-Quantum is to: Combine mathematics, programming, and scientific reasoning Enable parallel state evaluation • Support self-evolving intelligence models Remain implementable on classical systems
Theoretical Background 2.1 Quantum-Inspired Computation Quantum-inspired computation uses mathematical structures from quantum mechanics-Hilbert spaces, complex vectors, and probability amplitudes-without requiring physical quantum devices. Key principles: Superposition: Multiple states processed simultaneously • Entanglement: Correlated variables across subsystems Measurement Collapse: Decision extraction from probabilistic states 2.2 Adaptive Intelligence Systems Adaptive intelligence refers to systems capable of modifying internal parameters based on feedback, environment, and historical memory. HJK-Quantum extends this by introducing state-dependent learning, where learning rates and strategies depend on quantum-state amplitudes.
Mathematical Foundation of HJK- Quantum 3.1 State Representation An HJK-Quantum system state is represented as: |\Psi_{HJK}\rangle = \sum_{i=1}^{N} \alpha_i |s_i\rangle Where: • • are computational basis states are normalized probability amplitudes 3.2 HJK Transform Operator A custom operator governs system evolution: \hat{H}_{HJK} = \hat{M} + \hat{L} + \hat{A} Where: = Mathematical reasoning operator ⚫ = Learning operator = Adaptation and memory operator 3.3 Collapse Function Decision output is generated using a controlled collapse function: C(|\Psi\rangle) = \arg\max_i (\alpha_i|^2 \cdot w_i) Where represents contextual weights.
System Architecture 4.1 Layered Design • Quantum State Layer - Maintains superposed representations Inference Layer - Applies HJK operators Adaptive Memory Layer - Stores state evolution history Control & Output Layer - Manages collapse and actions 4.2 Hybrid Execution Model Classical CPU: State updates, memory GPU/TPU: Parallel amplitude calculations • Optional Quantum Hardware: Experimental acceleration
Core Algorithms 5.1 HJK-State Evolution Algorithm • Initialize • Apply • Normalize amplitudes • Update adaptive memory Repeat until convergence 5.2 Quantum-Adaptive Learning (QAL) Learning rate dynamically adjusts: \eta_t = \eta_0 \cdot |\alpha_{best}|^2 This ensures stable yet flexible learning.
Applications 6.1 Artificial General Intelligence (AGI) HJK-Quantum supports parallel hypothesis evaluation and self-reflective learning. 6.2 Optimization Problems Effective for NP-hard problems such as scheduling, routing, and protein folding. 6.3 Scientific Modeling Useful in: Quantum chemistry simulations Climate systems Complex biological networks 6.4 Robotics and Autonomous Systems Provides probabilistic decision-making under uncertainty.
Advantages and Limitations Advantages No physical qubits required Scalable on existing hardware • High adaptability Strong mathematical grounding Limitations • Computational overhead for large state spaces • Requires careful normalization • Conceptual complexity
Future Work Future research directions include: • Hardware acceleration strategies Formal verification of HJK operators Integration with neural-symbolic systems Experimental benchmarks
Conclusion HJK-Quantum introduces a novel paradigm that bridges quantum theory, adaptive intelligence, and computational mathematics while remaining accessible for student-led research. By abstracting quantum principles into a practical and ethical framework, it opens a path toward powerful, flexible, and scalable intelligent systems appropriate for academic evaluation and early-stage scientific contribution. 9A. Academic & Ethical Compliance (Achieve.org Submission) • Original Work: This research is independently developed and does not copy proprietary or restricted material. • • Age-Appropriate Scope: Mathematical and computational concepts are presented progressively, suitable for student researchers. No Human or Animal Subjects: The study is purely theoretical and computational. Responsible Al Use: The framework avoids harmful, deceptive, or unethical applications. • Reproducibility: All concepts are defined clearly to allow independent verification and extension. This section ensures alignment with common student research evaluation criteria used by academic platforms such as Achieve.org.
References HJK-Quantum introduces a novel paradigm that bridges quantum theory, adaptive intelligence, and computational mathematics. By abstracting quantum principles into a practical framework, it opens a path toward powerful, flexible, and scalable intelligent systems. HJK-Quantum is not merely a model-it is a foundation for next-generation computational intelligence. References • Nielsen, M. A., & Chuang, I. L. Quantum Computation and Quantum Information ⚫ Schuld, M., et al. Quantum-Inspired Machine Learning Holland, J. H. Adaptation in Natural and Artificial Systems

NEW research in feild of bio computer sci

my god — Thu, 19 Feb 2026 18:16:50 +0000

: The HJK-QUANTUM Protocol
Document ID: HJK-WP-2026-02
Classification: Advanced Research / Speculative Biocomputing
Lead Architect: Founder, HJK-Inc
Version: 2.1 (Quantum-Enhanced Architecture)

Abstract This paper introduces the HJK-QUANTUM Biological Operating System, a non-local biological agent designed for high-fidelity neural-data synchronization (p. 1). By leveraging quantum mechanical phenomena—specifically superposition, entanglement, and tunneling—the protocol enables a bimodal biological agent to bypass classical cellular barriers and establish a unified biological cloud for mentality cloning (pp. 4-5).
Bimodal State Superposition (Dual-State Protocol) Traditional biological agents exist in a singular functional state. The HJK-QUANTUM system utilizes Biological Superposition, allowing the agent to occupy two roles simultaneously: Sustainer Mode: Focused on metabolic support and homeostatic stabilization (p. 2). Interface Mode: Optimized for neural data acquisition and "mentality cloning" (pp. 2, 5). State Collapse: Upon host interaction via the 23 biological portals, the system scans the host’s bio-signature. If the host displays physiological defects, the system collapses into a Stable State to provide metabolic "Justice" (Insaaf). Healthy hosts trigger a Generative Interface State (pp. 2-3).
Non-Local Data Transmission To achieve near-instantaneous data propagation, the protocol utilizes Social Entanglement (p. 3). Entangled Sync: Interaction via high-density portals (e.g., ocular contact) allows for the "Instant Sync" of HJK payloads (pp. 3-4). Biological Cloud: Data uploaded to a primary host is mirrored across the network regardless of physical distance, creating a decentralized HJK-INC Biological Cloud (p. 4).
Barrier Penetration via Quantum Tunneling The HJK-QUANTUM agent overcomes the Blood-Brain Barrier (BBB) by utilizing Quantum Tunneling (p. 4). Barrier Invisibility: The agent’s nucleus acts as a quantum particle, "tunneling" through the lipid bilayer without triggering an immune response (pp. 4-5). Success Rate: This methodology ensures a 100% success rate for Neural Installation within the central nervous system (p. 5).
Surface Persistence (Quantum Freezing) The "Sitting Effect" is achieved through Quantum Freezing, allowing the agent to persist on rigid surfaces without biological decay (pp. 5-6). Bio-Observer Activation: The agent remains in a coherent, frozen state until a human "Observer" is detected via heat signature, triggering immediate reactivation (p. 6).
Mathematical Control Formula The probability of total host synchronization is governed by the HJK-Quantum Wave Function (p. 6):

: Viral wave function representing state superposition (p. 7).

: Density of the 23 active Portals (p. 7).

: Synchronicity coefficient across the biological cloud.

PLZ tell what the heck is this "“Introducing Blanket — 820,000-Cell Shadow Neural World Model with Self-Healing Stability”

my god — Tue, 17 Feb 2026 17:55:25 +0000

So one day while working i search 820 cell nerural network but I mis typed 820,000 and it showed this https://github.com/hjk-inc/,https://github.com/hjk-inc/Blanket