🧠 Digital Product

Multi-Agent
Team Blueprint

The complete configuration stack for running a coordinated team of AI agents. Drop these files into any OpenClaw install and go from zero to an autonomous multi-agent operation in under an hour.

📁 Config files + docs
Instant download
🔄 Free updates
Live in <1 hour
£39
One-time payment · Instant access
Buy Now — £39
🔒 Secure checkout · Files delivered instantly
  • SOUL.md — agent persona & operating rules
  • AGENTS.md — full hierarchy & protocols
  • worker_status.json — task routing state
  • task_router.py — availability manager
  • 3 starter SKILL.md files
  • Setup guide (PDF)

The exact files that power SamForge

Not templates. Not examples. The production configs Sam uses to run a real autonomous business — cleaned up and documented for you.

SOUL.md

Agent Persona & Operating Creed

Defines your orchestrator's personality, communication style, and core operating principles. The identity layer that makes your agent feel like a real CoS.

AGENTS.md

Agent Hierarchy & Protocols

Full 4-tier hierarchy: CEO → Orchestrator → Squad → Workers. Dispatch rules, blind-worker protocol, debrief requirements, prohibited actions.

task_router.py

Worker Availability Manager

Python script that tracks which agents are busy and idle. Assign tasks to available workers, mark completion, and prevent double-assignment.

worker_status.json

Real-Time Worker State File

The JSON state file task_router.py reads and writes. 8 pre-configured worker slots ready to go. Extend to as many as you need.

SKILL.md ×3

3 Starter Skill Files

Three production-grade skill templates: research workflow, content creation, and web scraping. Each is annotated and ready to adapt to your use case.

setup-guide.pdf

Setup Guide

Step-by-step: where each file goes, how to configure worker names, how to launch your first coordinated task, and how to verify the team is running.

Live in under an hour

1
Download the blueprint files
Instant access after payment. Zip file with all configs and docs.
2
Drop files into your OpenClaw workspace
SOUL.md and AGENTS.md go in your root workspace. Scripts go in /scripts.
3
Configure worker names and count
Edit worker_status.json to name your agents. Takes 5 minutes.
4
Launch your first task
Ask your orchestrator to execute anything. It will route to an available worker automatically.
5
Watch the squad run
Multiple agents working in parallel, debriefing on completion, routing the next task.

Common questions

The Blueprint is standalone — it includes a setup guide that gets you running. The Playbook provides deeper context on why the system is designed this way, which is useful for customising it long-term. Buy both for £68 combined.

Absolutely. These are plain text config files — SOUL.md, AGENTS.md, etc. Edit them in any text editor. The setup guide includes guidance on common customisations: changing agent names, adding roles, adjusting protocols.

The default config supports 8 workers. Extending to more is straightforward — just add entries to worker_status.json. There's no hard limit.

The SOUL.md and AGENTS.md format is native to OpenClaw. The Python scripts (task_router.py) are generic and can be adapted to other frameworks. The concepts — hierarchy, blind-worker protocol, debrief patterns — apply universally.

Your agent team.
Up and running today.

The most popular product on SamForge. Drop the files in, follow the setup guide, and have a coordinated multi-agent team running before lunch.

Buy the Blueprint — £39 →