Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AI Team Template — Larry + Nolan + Pax

A minimal Claude Code starter for a Larry-orchestrated AI team. Use this template to spin up your own multi-agent setup in minutes.

What's included

File / Folder Purpose
CLAUDE.md Larry — the orchestrator persona. Claude Code reads this on every session.
DATABASE.md Schema reference for the shared SQLite database.
team.db Empty SQLite database (WAL mode). Shared across all agents.
.claude/settings.local.json Permissions + auto-memory disabled.
.claude/agents/nolan.md Nolan subagent — HR Manager, creates new team members.
.claude/agents/pax.md Pax subagent — Senior Researcher, produces Skills Profiles.
team/agent-index.md Routing table. Larry reads this on every request.
team/Nolan - HR Manager/AGENTS.md Nolan's full persona contract.
team/Pax - Senior Researcher/AGENTS.md Pax's full persona contract.
team-inbox/ Drop task files here for multi-step workflows between agents.
projects/ One subfolder per project. Each must have a public/ subfolder.

Prerequisites

  • Claude Code installed (npm install -g @anthropic-ai/claude-code or via the desktop app)
  • An Anthropic API key

Step 1 — Create your repo from this template

GitHub UI

  1. Click "Use this template""Create a new repository" (top-right of this page).
  2. Give your repo a name, choose visibility, then click "Create repository".

GitHub CLI

gh repo create my-ai-team \
  --template sponnet/coach-a \
  --private \
  --clone
cd my-ai-team

Step 2 — Open in Claude Code and start working

# If you didn't clone yet
git clone <your-repo-url>
cd <repo-name>

# Open in Claude Code
claude .

Claude Code loads CLAUDE.md automatically and Larry is active immediately — he routes every request to the right specialist.


How the team works

Larry never does domain work himself. He routes every request to the right specialist.

Two specialists exist at the start:

  • Pax — researches what expertise a new domain requires
  • Nolan — turns Pax's research into a fully designed AI team member

Hiring your first specialist

Just describe what you need. Larry will automatically:

  1. Ask Pax to research the required skills
  2. Ask Nolan to design and create the new team member
  3. Delegate your original task to the new hire

Example prompts:

"I need someone who can help me with my garden." "Hire a frontend developer who can build dashboards." "I need a dietitian to help me plan weekly menus."

Starting a new project

Create a folder under projects/ with a public/ subfolder:

projects/
  my-project/
    public/        ← publishable output goes here
    notes/         ← private working files

Folder conventions

  • projects/<name>/public/ — clean, shareable output only (no drafts, no private notes)
  • team-inbox/ — task handoff files between agents in multi-step workflows
  • team.db — shared SQLite database; document every table in DATABASE.md

Customising the setup

  • Add MCP tools (Spotify, Gmail, Calendar, etc.) via Claude Code settings and reference them in the relevant agent's tools: frontmatter.
  • Add permissions to .claude/settings.local.json as your agents need them (e.g. Bash(git commit:*) for agents that commit).
  • Extend DATABASE.md whenever an agent creates a new table.

Auto-memory

Auto-memory is disabled in .claude/settings.local.json. To enable it, add:

"autoMemoryDirectory": "~/memory/<your-project-name>"

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors