Feature Request: Enable Agent Teams on all plans

Resolved 💬 7 comments Opened Feb 12, 2026 by brandonmanson Closed Apr 8, 2026

Request

Make Agent Teams feature (TeammateTool, SendMessage, spawnTeam) available on all plans, not just enterprise.

Current State

System prompt shows:

"Note: The 'Agent Teams' feature (TeammateTool, SendMessage, spawnTeam) is not available on this plan."

The feature exists but is gated behind plan tiers.

Use Cases

Complex multi-agent workflows need peer-to-peer coordination:

1. Parallel code review workflow:

  • Planning agent creates task breakdown
  • Multiple implementer agents work in parallel
  • Each implementer hands off to dedicated reviewer
  • Reviewers communicate findings to each other
  • Final integration agent synthesizes results

2. Specialist agent collaboration:

  • Frontend agent ↔ Backend agent coordinate API contracts
  • Test agent validates both implementations
  • Agents negotiate design decisions in real-time

3. Dynamic task delegation:

  • Lead agent assesses complexity
  • Spawns specialist sub-agents based on needs
  • Sub-agents report back with results
  • Lead agent adapts strategy based on feedback

Current Limitations

Without Agent Teams:

  • ❌ Main agent must orchestrate everything
  • ❌ No peer-to-peer communication
  • ❌ Can't build self-organizing agent systems
  • ❌ Limited to hub-and-spoke architecture

Current workaround: Spawn agents via Task tool, coordinate through shared state (files, task lists). But this is:

  • Brittle (race conditions, state conflicts)
  • Inefficient (polling for changes)
  • Limited (no direct messaging)

Benefits of Enabling for All Users

  1. Unlock advanced workflows - Users can build sophisticated multi-agent systems
  2. Better UX - Agents that coordinate themselves vs manual orchestration
  3. Showcase capabilities - Demonstrates Claude's multi-agent potential
  4. Community innovation - Users will discover novel coordination patterns

Alternative

If full Agent Teams can't be enabled for all plans, provide lightweight coordination primitives:

  • AgentMessage tool for peer-to-peer messaging
  • Shared context/state management
  • Event bus for agent coordination

Why This Matters

AI-native development needs AI-native coordination. Current limitation forces users back to traditional orchestration patterns instead of letting agents self-organize.

Environment

  • Claude Code CLI version: 2.1.32
  • Platform: macOS (Darwin 23.6.0)
  • Plan: [Pro/Team - whatever the user has]

View original on GitHub ↗

This issue has 7 comments on GitHub. Read the full discussion on GitHub ↗