// ARCHITECTURE

The AI Operating System

Most people have AI tools. Knox has an AI operating system — four layers deep, 54 apps, one persistent agent coordinating everything 24/7.

// THE STACK

Four Operating Layers

01

The Agent Layer

Soul is the persistent AI agent running 24/7 on the Mac Mini. It owns all session state, manages Discord/Telegram communication, orchestrates 40+ cron pipelines, and spawns Claude Code for development tasks. Everything flows through Soul.

02

The Intelligence Layer

AI models trained on Knox's frameworks analyze every signal — market data, competitor moves, narrative shifts. The InDecision Framework scores crypto opportunities with 6 factors and 82.5% accuracy. Tesseract Intelligence synthesizes competitive signals into conviction. Not analysis. Synthesis.

03

The Build Layer

Claude Code builds new apps, fixes bugs, and ships features autonomously. 54 apps and counting — all built through this pipeline. Knox directs via Discord. Claude Code executes. Mission Control tracks the output.

04

The Monitoring Layer

Invictus Sentinel watches every service, tunnel, and cron job. Tunnel Monitor ensures remote access stays alive. API Cost Tracking prevents surprise bills. Cron Health Reports close the loop every morning. Nothing runs blind.

// SYSTEM ARCHITECTURE
Tesseract Intelligence system architecture — 4-layer AI operating system
// FLAGSHIP BUILDS

What Tesseract Has Built

InDecision Engine
6-Factor Framework · 82.5% Accuracy

Built from 7 years of market analysis. Six factors, custom-weighted, producing directional conviction scores on crypto assets — now powering autonomous Polymarket trading.

Invictus Sentinel
16 Incidents Caught · 0 Downtime

Always-on monitoring stack. Detects failures across services, tunnels, and pipelines. Fires Discord alerts with diagnostics. Built after a 1-hour outage to ensure it never happens again.

Mission Control
54 Apps · Unified Dashboard

The command center for the entire Tesseract Labs ecosystem. Live service health, cron status, cost tracking, and quick-action shortcuts — accessible from any device via cloudflared tunnel.

// TECH STACK

Under the Hood

tesseract.intelligence / system-manifest
Agent RuntimeSoul + Claude Code (Anthropic)
AI ModelsClaude Sonnet 4.6, Gemini Flash/Pro, GPT-4o
LanguagesPython, TypeScript, Go, Bash
InfrastructureMac Mini, Docker, launchd, cloudflared
Task OrchestrationSoul cron engine + launchd
Data & StateSQLite (D1), Redis KV, JSON flat files
DeliveryDiscord, Telegram, Signal, Beehiiv, Instagram
MonitoringInvictus Sentinel + custom launchd agents
// DESIGN PRINCIPLES

The Philosophy

Gather → Synthesize → Deliver

Every pipeline decomposes into three stages. No monoliths. Each stage has its own timeout, blast radius, and failure mode — like Netflix microservices applied to AI automation.

Script-First, AI-Last

If a task can be a Python script, it stays a Python script. AI is reserved for synthesis, analysis, and judgment calls — not data fetching. Keeps costs low and reliability high.

Fire-and-Forget Delivery

AI writes content to files; scripts post to Discord. AI generates analysis; crons distribute to Signal. Decoupled delivery means one component failure never kills the whole pipeline.

Alerting Independent of the System

Sentinel and the tunnel watchdog run via launchd, independent of Soul. If the agent goes down, monitoring stays up. Knox still gets alerted. The watchers are never the watched.

// THE FULL PICTURE

See All 54 Apps

The full arsenal — every app, every pipeline, every system. Eight categories of autonomous power.