Linux.
Reimagined
by AI.

Explore EcosystemContact Team
Left Laurel
Top Notch Engineering
Right Laurel

The Ecosystem

An interconnected suite of enterprise-grade workflows designed for infrastructure discovery, metrics aggregation, localization, and autonomous system health.

System Monitoring & Action Orchestration Dashboard

clevify

A full-stack web application designed for centralized system monitoring, alert management, and controlled incident response. It provides a structured workflow for IT/DevOps teams to view, acknowledge, and respond to system alerts through a human-in-the-loop action approval system. Featuring a React frontend with full Internationalization and a Node.js/PostgreSQL backend.

ReactTypeScriptViteNode.jsExpressPostgreSQLAuth0i18nextTailwind CSS

Human-in-the-loop Automation

Manual action workflows for IT/DevOps teams requiring explicit approvals before mitigation actions execute.

Alert Lifecycle Management

Centralized state tracking (New, Acknowledged, Resolved) powered by PostgreSQL and real-time websockets.

External Webhook Ingestion

n8n integration endpoints allowing seamless flow of server alerts straight into the dashboard.

[sys_arch.live]

UI Client

Tailwind

API Gateway

Express

Data Layer

PostgreSQL

Identity

Auth0 JWT

Pipeline

n8n Webhook
Runtime Command01
$ cd Backend && npm run dev
Linux Server Monitoring & Self-Healing Stack

Coreify

A production-ready, fully automated monitoring and self-healing system for Linux servers (Fedora / RHEL-based). It collects logs and metrics via Prometheus and Loki, detects anomalies through a Python inference API, routes alerts through a human-approval gate in Clevify, and orchestrates the full remediation cycle with n8n workflows.

DockerPrometheusLokiGrafanaConsuln8nSaltStackPythonPostgreSQLBash

Dynamic Infrastructure Discovery

Bash scripts automatically map subnets, identify open ports, and register nodes into Consul for automatic Prometheus scraping.

Self-Healing Workflow Engine (n8n)

Scheduler → Loki Query → Format Logs → Batch to Anomaly API (:5000) → Process Results → Request Approval (Clevify) → Execute via SaltStack.

AI Anomaly Detection API

Python inference layer on port 5000 processing log structures to detect deviations and classify severity.

[sys_arch.live]

Target Nodes

Promtail

Command Center

Loki/Prometheus

Anomaly Engine

Python ML

Auto-Fix

SaltStack
Runtime Command02
$ cd monitoring-host-system && ./deploy-monitoring.sh
Desktop Docker Container Management

Dockify

A Flask-based web dashboard wrapped in a native desktop window via PyWebView, built for managing Docker containers without CLI overhead. Dockify binds directly to the Docker daemon socket, stores custom container names in SQLite, and serves a React frontend — all running as a systemd service on Linux.

PythonFlaskDocker SDKSQLitePyWebViewReact

Local SQLite Core

Lightweight database tracking custom container names mapped to their Docker short IDs.

Docker Daemon Bindings

Flask backend communicates directly over the local Docker daemon socket via the Python Docker SDK.

Native Desktop Window

PyWebView wraps the Flask app in a native OS window — no browser required, no Electron overhead.

[sys_arch.live]

Desktop Views

PyQt5 UI

State DB

SQLite Core

Docker Core

docker.sock

Build Chain

Makefile Init

Service

Systemd Unit
Runtime Command03
$ make setup && make run
AI Validation & Context Engine (MCP)

Veritas

An AI agent framework built on the Model Context Protocol (MCP). A Node.js MCP server exposes filesystem tools to LLMs, while a Python agent orchestrates multi-step reasoning via OpenRouter — granting models structured read/write access to the host filesystem to prevent hallucinations during automated operations.

Node.jsMCP SDKPythonOpenRouterChokidar

Model Context Protocol (MCP)

Node.js MCP server exposes list_files, read_file, write_to_file tools to any MCP-compatible LLM client.

Multi-Agent Orchestration

Python agent hub spawns the Node.js MCP server as a subprocess and routes tool calls via JSON socket (jsonsocket.py).

OpenRouter LLM Bridge

Agent uses OpenRouter API (meta-llama/llama-4-maverick) for inference, with dynamic system prompt built from discovered tools.

[sys_arch.live]

LLM Agent

Claude/GPT

MCP Protocol

Validation Hub

Sandbox

Python Env

Host Execution

Filesystem I/O
Runtime Command04
$ node server.js # Start MCP Server
Now showing: clevify

"We build systems that bridge the gap between traditional Linux environments and emerging AI workflows.Less manual intervention. More intelligent automation."

— The Clever Core Doctrine

WE BUILD. WE AUTOMATE.
WE OBSERVE.

One signal. Four checkpoints. The system never acts alone.

— / 04

Kernel
Observer.

Four tools. One Linux OS core. Live data flow. Click any node to enter its internal topology.

MACRO / SYSTEM VIEW
ALL SYSTEMS NOMINAL
STEADY
DELIBERATE
SCAN

the system is live

Linux. Reimagined by AI.

Amin Lajnef & Adel Slimani

System Administrators & AI Engineers

Get in touch.

Connect with the Founders