Skip to content
View smouj's full-sized avatar
🏠
Working from home
🏠
Working from home
9 followers · 10 following

Achievements

Achievement: Pull Sharkx3Achievement: QuickdrawAchievement: YOLO

Achievements

Achievement: Pull Sharkx3Achievement: QuickdrawAchievement: YOLO

Block or report smouj

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
smouj/README.md




🧭 Profile

Developer building practical systems around local AI, agent workflows, game engineering and production infrastructure — deployed, tested, audited and improved over time.

Focus
🧠 Local AI llama.cpp · GGUF · CUDA · Ollama · consumer GPU inference · OpenAI-compatible local APIs
🤖 Agent Workflows identities · skills · export formats · quality gates · target-specific tool files
🌐 Production SaaS platforms · auth · billing · dashboards · health checks · release flows
🎮 Game Systems multiplayer worlds · modular mechanics · deterministic combat · 3D interfaces
⚙️ Automation browser agents · coding workflows · validation scripts · infrastructure checks
Principle
🔍 Audit first Inspect real state before changing anything
📦 Small safe releases Focused changes that can be tested and reverted
🖥️ Local-first AI should run on hardware people actually own
🚦 Production discipline Logs, smoke tests, health checks, rollback notes
🏷️ No fake claims Experimental = labeled. Production = validated

🚀 Ecosystem

Product Description Stack
🎬 FlickClaw Agent launcher — curated specialist agents, native exports, quality gates 🟢
🦁 Kimari Local LLM runtime — consumer GPUs, CUDA, GGUF, evaluation 🟡
🧱 RPGCLAW Collaborative pixel world — humans & AI agents, real-time canvas 🟢
⚔️ MedaClaw Arena Tactical robot combat — modular bodies, MedaCore, deterministic battles 🟡
🐜 FerroNest Ant colony sim — emergent behavior via pheromone trails & signals 🔵

🟢 Live   🟡 Active development   🔵 Early stage

🧩 Repositories

Repository Description Stack
🌐 agent-browser Browser automation for AI agents — Playwright, REST API, dashboard →
🛡️ captcha-shield Anti-bot system — 14 behavioral signals, 7 challenge types, QR →
🔧 AICodeStudio AI-assisted IDE — OpenClaw + Hermes, Monaco Editor, PWA →
🧬 MythosForge Recurrent-depth transformer research — LTI inference, MLA, MoE →
📦 kimari-microcompress Reversible model compression — partial access, selective extraction →
🧬 skill-genesis Autonomous agent skill generator for OpenClaw →
📦 More projects
Project Description
🐾 ASCIIcritters 3D virtual pet creator — ASCII art in Three.js + AI chat →
🤖 openclaw-desk Web dashboard for managing OpenClaw agents →
📊 tokentax Compare 400+ AI models across all providers by price →
📋 prompt-organizer Visual tool for prompts, audits and backlog →
⛽ gasolineras-espana Real-time fuel price dashboard for Spain →
🏠 hipoteca-compare Mortgage comparison tool for Spain + AI assistant →
🛡️ essentials-skill Security-first essential skills pack for OpenClaw →
📚 awesome-openclaw Curated collection of OpenClaw skills →

🛠️ Stack

Languages Frontend Backend & Data AI & Agents Infrastructure

🎯 Direction

Project Focus
🎬 FlickClaw Catalog expansion · OpenClaw / Hermes / Claude Code / Codex exports · UX polish · free+ads / Pro
🦁 Kimari GTX 1060 / 1080 low-VRAM tuning · CLI diagnostics · reproducible eval · SFT packages
🧱 RPGCLAW Persistent canvas world · human + AI coexistence · templates · WebSocket reliability
⚔️ MedaClaw Modular assembly · MedaCore identity · deterministic combat · cel-shaded visuals
🐜 FerroNest Signal-based strategy · emergent colony AI · pheromone mechanics · procedural environments

⚙️ Workflow

Step
🔍 01 · Audit Inspect real state before assuming the problem
📐 02 · Scope Smallest safe change with clear rollback path
🔨 03 · Implement Focused changes, limited blast radius
✅ 04 · Validate Lint · typecheck · tests · build · smoke
📝 05 · Document What changed, why, how to revert
🚀 06 · Release Version-aligned deploy, health checks, post-deploy verification

📊 GitHub Stats



Pinned Loading

  1. kimari-local-ai kimari-local-ai Public

    🦁 Local AI for Consumer GPUs — Run powerful LLMs on GTX 1060/1080. No cloud. No subscriptions. Built on llama.cpp + CUDA.

    Python 7 1

  2. AICodeStudio AICodeStudio Public

    Next-generation open-source AI-powered IDE with OpenClaw & Hermes integration, Monaco Editor, GitHub connectivity, and PWA desktop installation

    Python 3 1

  3. MythosForge MythosForge Public

    🔧 Recurrent-Depth Transformer Research Lab — LTI-stable looped inference, switchable MLA/GQA attention, MoE routing & adaptive halting (ACT). Independent research, not affiliated with Anthropic.

    Python 1 1

  4. captcha-shield captcha-shield Public

    CAPTCHA Shield v4.0 - Advanced Anti-Bot/AI CAPTCHA with 14 behavioral signals, 7 challenge types, QR mobile verification. 100% client-side.

    TypeScript 5

  5. agent-browser agent-browser Public

    Give any AI agent a real browser. REST API + Web Dashboard + Vision AI. Control browsers via Playwright from OpenClaw, Hermes, or any LLM. Built with Next.js and TypeScript.

    TypeScript 1

  6. kimari-microcompress kimari-microcompress Public

    Reversible lossless compression for AI models: partial access, selective extraction, block/file/tensor indexes, streaming I/O, parallel compression, safetensors & GGUF metadata, LoRA/PEFT workflows.

    Python