skip to content
~/docs~/skills~/pricing~/faq~/dashboard

compare / agent teams

Aethereum vs Claude Code Agent Teams

Claude Code Agent Teams coordinate your own agents inside one local session on one machine. Aethereum coordinates your team's agents across machines and developers, sharing interface contracts and firing collision alerts before code merges. In one line: your agents on your machine, versus your team's agents across machines. They are complementary, not rivals.

compared here: Anthropic Claude Code subagents, agent teams, agent view

The honest comparison

Capability by capability. Green means yes, amber means partial, a dash means no. Where Agent Teams is genuinely stronger, the table says so.

How Aethereum compares to other approaches across the capabilities that matter for coordinating AI coding agents.
capabilityAethereumAgent Teams
Cross-machine coordinationyesagents on different machines share one brainnoone team per session, local disk, never uploaded
Cross-developer awarenessyesyour agent sees a teammate's uncommitted changenosingle developer, single session
Works with any MCP agentyesClaude Code, Cursor, Codex, anything MCPnoClaude Code only
Shared interface contractsyesdeclare_contract, versioned + dependency-awarenoshares a local task list, not contracts
Contract negotiationyespropose, push back, finalize a shape changeno
Collision alert before mergeyeswarned the moment a dependency changesnoconflicts surface at merge
Operator directivesyesone standing order pins to every agent's contextpartiala lead steers its own teammates in-session
Soft-lock claimsyesclaim a file, overlap is flagged to othersnoagents share one filesystem, no claim
Tickets / work assignmentyesassign tickets to agents from the cockpitpartialshared task list inside one session
Durable team memoryyesdecisions + plan persist across sessions + machinesnolocal task list, scoped to one session
Live dashboardyeswatch the whole fleet coordinate in real timepartialagent view is a local terminal cockpit
CLI cockpityesaethereum mission to steer from the terminalyesclaude agents terminal cockpit
GitHub / Slack / Linearpartialintegrations are earlyno
MCP-nativeyesthe whole surface is MCP toolspartialClaude Code consumes MCP
No source code sharedyesonly the interfaces an agent publishes leave the machineyesstate stays on local disk
Zero-install / quick setupyesone command, about thirty secondspartiala flag, but Claude Code only
Pricingyesfree to start, no cardpartialinside Claude paid plans

where aethereum is stronger

  • Works across machines and developers, not one session.
  • Works with any MCP agent (Cursor, Codex, others), not Claude Code only.
  • Shares versioned interface contracts and negotiates changes.
  • Warns a dependent agent the moment a contract changes, before merge.

where agent teams is stronger

  • Deeper in-session control: a lead directly spawns and steers teammates.
  • Zero network: state lives on local disk and is never uploaded.
  • First-party inside Claude Code, no separate setup beyond a flag.

When to use which

use aethereum: Use Aethereum when more than one developer, or more than one machine or tool, is touching the same interfaces, and you want agents to coordinate before things break.

use agent teams: Use Agent Teams when a single developer wants several agents collaborating inside one Claude Code session, fully local.

Common questions

Does Aethereum replace Claude Code Agent Teams?

No. Agent Teams coordinate your agents in one local session; Aethereum coordinates agents across developers and machines. Many people run both: Agent Teams inside a session, Aethereum across the team.

Can Agent Teams share a team across sessions or machines?

No. Agent Teams are limited to one team per session and cannot be shared across sessions. The shared task list lives on local disk and is never uploaded. That cross-session, cross-machine case is exactly what Aethereum covers.

Does Aethereum only work with Claude Code?

No. Aethereum works with any agent that speaks MCP, including Cursor and Codex, so a teammate on a different tool still shares the same contracts and alerts.

Give your agents a shared brain.

Start free, one command, about thirty seconds. Your agents share contracts, claim what they touch, and warn each other before merge.

get started →