Allie K. Miller Just Dropped a Claude Agent Teams Tutorial (And It's Actually Good)

📜 The Signal

Allie K. Miller — former Amazon AI lead, 2M+ audience — just published a breakdown of Claude Agent Teams in her "AI with ALLIE" newsletter today. This matters because she's explaining bleeding-edge Claude Code features to a non-technical audience, and she's doing it well.

Here's what she covered and what YOU should know as a pirate who's already in the water.

———

⚓ The Three Ways to Run Multiple Agents

Allie breaks down three distinct approaches. If you've been using Claude Code for a while, you probably already use #1 and #2 — but #3 is the new hotness.

Parallel Sub-agents — Ask Claude to spin up multiple agents in one session. Best for tasks that can be split but don't need to talk to each other.
Multiple Terminals — Open separate terminal windows, each running Claude Code. Best when YOU are the coordinator and tasks are completely independent.
Agent Teams (NEW) — Agents collaborate, share context, and check in with each other. Best for complex projects where agents need to coordinate.

———

⚓ What Agent Teams Actually Are

Allie uses a perfect analogy: four families planning a Sedona retreat. One parent handles Airbnbs, one handles hikes, one handles restaurants — and they coordinate through a group chat and shared doc.

Agent Teams work the same way:

Team Lead — the orchestrator who divvies out tasks
Teammates — separate agents each owning a piece of the work
Task List — running list of things to get done
Mailbox — how agents communicate with each other (DMs are cheaper than group messages)

Key detail she flags: agents do NOT see the full conversation history. They only get the "spawn prompt" from the team lead. So front-load your context.

———

⚓ How to Enable Agent Teams

It's one setting. Add this to your ~/.claude/settings.json:

"env": { "CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1" }

If you already have a settings file (and if you're in this group, you almost certainly do), just add the CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS line inside your existing "env" block.

Restart Claude Code. Done.

———

⚓ Her Best Example Prompt

This one is worth stealing and adapting:

"Create an agent team to stress-test my Q1 growth strategy. One teammate as a skeptical CFO poking holes in financial assumptions, one as a competitor analyst identifying counter-moves, one as an angry exec finding reasons this makes their life worse, and one as an operations lead flagging what would break in execution. Have them debate each other. The team lead should synthesize their critiques into a single brutal-but-constructive memo with prioritized action items."

That's a legitimate use case — adversarial review through multiple perspectives. You could adapt this for code review, security audits, content strategy, or any scenario where you want diverse viewpoints.

———

⚓ Pro Tips She Shares

You don't have to specify agent count — Claude can decide how many teammates and what roles
Use planning mode for risky tasks — tell Claude "I need to provide plan approval before these agents make any changes"
Message individual agents — Shift+Up/Down to select a specific teammate and talk to them directly
Shut down one agent — just say "Ask the [role] teammate to shut down"
Clean up when done — ask the team lead to "clean up the team" to spin down shared resources
Color-code your terminals — Shell > Edit Background Color on Mac (real ones know)

———

☠️ Watch Out

This is experimental — expect bugs, misunderstandings, and rough edges
Token burn is real — each teammate is its own Claude session. Allie says you probably need a Max plan, and it "can quickly add up to hundreds of dollars a day" on API
Don't overthink simple tasks — her words: "Don't ask a Michelin chef to make you toast." If one agent can handle it, skip the team
No nested teams — no teams within teams. One team lead, one team at a time

———

🗝️ Key Takeaways

Agent Teams let agents coordinate — unlike parallel sub-agents, they share context and communicate
One setting to enable — add the env variable, restart, done
Best for complex, multi-angle work — code review from different perspectives, multi-part projects, adversarial analysis
Watch your token spend — this feature multiplies your usage fast
Allie is pushing Claude Code to non-technical users — her 2M+ audience is discovering what we already use daily

———

🗺️ Source