Field Note6m read
Knox Is Building Itself
The moment an AI system starts planning its own next evolution is the moment you stop being a developer and start being a co-founder.
February 24, 2026Read →
SYSTEM ONLINE · 24/7
Meet Knox
A self-evolving AI agent system built from scratch — not downloaded, not configured, not purchased. Built, instruction by instruction, over seven months.
It tells me what to build. I approve it. The rest takes care of itself.
80%
Built by Claude Code
5
Connected AI Systems
7mo
Running Continuously
// THE JOURNEY
From notification layer to autonomous system.
Month 1
The comms layer
OpenClaw on a Mac Mini. Discord routing. Cron jobs. A smarter notification system.
Month 2
The builder arrives
Claude Code starts writing 80% of the implementation. Role shifts from developer to director.
Month 3
Memory system
lessons.md per project. CLAUDE.md globally. Mistakes stop repeating. Compounding begins.
Month 4
The trading mind
InDecision connects to Tesseract. Trade outcomes feed post-mortem analysis. Model starts learning.
Month 5
Asana goes autonomous
Tasks get created by OpenClaw, Claude Code, and Tesseract — not just by me. Knox manages its own backlog.
Month 7
God mode
Knox files its own bug reports. Assigns fixes to Claude Code. Runs retrospectives on itself. Still running.
// THE STACK
One system. Five minds.
Each component does exactly one thing well. The system-level behavior emerges from the interfaces between them.
OpenClaw
Always-on nerve center. Runs crons, handles Discord comms when AFK, routes tasks, spawns coding agents.
Connects to:Everything
Claude Code
The builder. Scopes work, writes implementation, opens PRs, inherits all project lessons from prior sessions.
Connects to:GitHub, Asana, CLAUDE.md
Tesseract Intelligence
Deep reasoning engine. Post-mortems, trade analysis, architectural decisions. Thinks — doesn't execute.
Connects to:InDecision, OpenClaw
InDecision
The trading mind. Six-factor crypto bias model. Learns from every trade via Tesseract post-mortem analysis.
Connects to:Tesseract, PolyMarket
Asana
Knox's external memory. Tasks flow in from OpenClaw, Claude Code, and Tesseract — not just from me.
Connects to:OpenClaw, Claude Code
// OPERATING PRINCIPLES
The rules that make it work.
Approver, not implementor
Every PR waits for my review. Every Asana task is visible. But the scoping, building, debugging, and deployment — that's Knox's problem.
Retrospectives compound
Every mistake becomes a rule. Every rule prevents the next mistake. The system doesn't plateau — it gets relentlessly less wrong.
Clean component boundaries
OpenClaw routes. Claude Code builds. Tesseract reasons. InDecision trades. Asana tracks. No component does another's job.
// FIELD NOTES
Inside the build. No filter.
Field Note10m read
From Discord Bot to God Mode: The Knox Story
This isn't a story about building AI tools. It's a story about what happens when the tools start building themselves — and you realize your job has fundamentally changed.
February 24, 2026Read →
Field Note8m read
The God Squad: Inside the Architecture
OpenClaw handles comms. Tesseract does the thinking. InDecision trades. Claude Code builds. Asana tracks. One system, five minds.
February 23, 2026Read →
Field Note5m read
Why Retrospectives Are the Only Thing That Matter
You can give an AI perfect tools and still get mediocre output. The difference between a good agent and a great one is structured feedback over time.
February 22, 2026Read →
Field Note7m read
Teaching the Trading Mind: How InDecision Learns from Every Trade
Most bots optimize for wins. Knox's trading mind learns from losses too — using Tesseract's deep reasoning to find the real failure mode.
February 21, 2026Read →
// STILL RUNNING
Knox is still being written.
Every week it ships something new. Every mistake becomes a rule. Every rule makes the next session better.