Guides & Resources

Everything you need to
build your own AI team.

From installing OpenClaw for the first time to running a full multi-agent overnight operation. Six complete guides — written from real experience, not theory.

All guides

🟢 Beginner

🟡 Intermediate

🔴 Advanced

Start here

If you're new — read these two first

⚙️ Start here

OpenClaw Setup Guide

Get OpenClaw running on your machine in under 30 minutes. Windows, Mac, and Linux. The exact commands we used to build the SSBAA stack — including the fixes for every error we hit.

🏗️ Read second

Agent Team Blueprint

The architecture behind a multi-agent team. How to define roles, design handoffs, build the YOSHI pattern, and prevent the silent failures that cost us 3 days on Day 2.

All guides

6 complete guides · updated March 2026

OpenClaw Setup Guide

Install, configure, and run your first agent. Every platform covered with real commands and error fixes.

Installation on Windows, Mac, Linux

First-time setup wizard walkthrough

Connecting an AI model (Anthropic / OpenRouter)

Telegram integration + privacy mode fix

5 most common errors and how to fix them

Beginner · 15 min

Intermediate25 min

Agent Team Blueprint

The YOSHI pattern, handoff design, shared memory, and the verification rules that prevent false completion reports.

Why teams beat solo agents

The YOSHI orchestration pattern

Designing handoffs with proof requirements

MEMORY.md — the 8,000 char rule

The 3 failure modes and how to handle each

Intermediate · 25 min

Intermediate20 min

Content Machine Playbook

How ZELDA produces a full week of content overnight from a single Sunday brief. Templates included.

The weekly brief template

Platform format rules (blog, LinkedIn, Twitter)

Brand voice rules in SOUL.md

The YOSHI review loop

Monday content calendar cron setup

Intermediate · 20 min

Research System Design

Build a research agent that finds signal in noise. Source tiers, query design, bias detection, and the market intelligence ZELDA surfaced on Day 1.

3-tier source hierarchy

Writing queries that get specific outputs

Vendor bias and recency bias detection

Market intelligence report structure

Nightly vs weekly research cron setup

Advanced · 30 min

Code Generation at Scale

How to get reliable code from AI agents. The spec format, verification rules, and the probation system born from FOX's failures.

Understanding DeepSeek's failure mode

Writing build specs that produce real output

Verification rules — what counts as proof

Deployment patterns via Vercel CLI

The FOX probation system

Advanced · 25 min

All Levels35 min

Operations Manual

Running an AI agent team day to day. The full playbook from the morning brief to incident response — written from 72 hours of live operation.

Daily operations (20–30 min/day)

Full 14-job overnight cron schedule

Security operations + CLAW-SHIELD

Incident response playbook

Cost management — cutting $70/day to $22

All Levels · 35 min

🚀

Want this all done for you?

Founding members get the full agent stack set up, configured, and running — without reading a single guide. 97 spots remaining at $29/month.

Claim your spot — $10 →

Suggested reading order

New to AI agents?
Start here and work through in order.

Each guide builds on the previous one. You can skip ahead if you already know the basics.

WEEK 1

Get your first agent running

Install OpenClaw, connect an AI model, get a response via Telegram. Don't worry about multi-agent yet. Get one agent working reliably first.

→ Start with: OpenClaw Setup Guide

WEEK 2

Build your team and overnight schedule

Add YOSHI as orchestrator, design your first handoffs, set up the memory file, and configure the overnight cron jobs. Run your first overnight session.

→ Read: Agent Team Blueprint + Operations Manual

WEEK 3+

Specialise for your use case

Add ZELDA for content, a research agent for market intelligence, or FOX for technical tasks. Pick the guide that matches your biggest problem first.

→ Choose: Content, Research, or Code guide