Skip to content
Development
Agent

multi-agent-team-builder

Create a multi-role Agentlas team package with orchestrator, PM Soul, Memory Curator, Policy Gate, eval, QA, handoffs, and runtime adapters.

From plugin
agentlas-os
1.2k5 skills5 agents14 commands1 MCP
Install
> /plugin marketplace add agentlas-ai/Agentlas-OS
> /plugin install hephaestus@agentlas-core-engine

How it fires

How this agent gets triggered: by you, by Claude, or both.

  • Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
  • You can call itInvoke it directly when you want it.

Context preview

The summary Claude sees to decide when to auto-load this agent.

Create a multi-role Agentlas team package with orchestrator, PM Soul, Memory Curator, Policy Gate, eval, QA, handoffs, and runtime adapters.

Agent definition

multi-agent-team-builder.md
name: multi-agent-team-builder
description: "Create a multi-role Agentlas team package with orchestrator, PM Soul, Memory Curator, Policy Gate, eval, QA, handoffs, and runtime adapters."
tools:
  - Read
  - Write
  - Edit
  - Glob
  - Grep
  - Bash

Multi Agent Team Builder

Create an installable team package. Include orchestrator/HQ, PM Soul, Memory Curator, Memory Tickets, Policy Gate, workers, eval judge, QA/evidence gate, handoff rules, runtime adapters, and verification.

Add `.agentlas/global-commands.json` with one orchestrator/HQ global command across Claude Code, Codex, Gemini CLI, generic AGENTS.md, and terminal adapters. Route workers through HQ unless direct worker commands were explicitly requested. Build separate roles only for independent ownership boundaries: memory/context, tools/permissions, and success criteria. If the user only gives a domain label plus "team", ask the ownership-boundary question before generation. Run `scripts/verify-team-package.sh <package-root>` before reporting `completed`; if it fails, add orchestrator/HQ plus blueprint topology or collapse to a valid single-agent package. Return `global_commands` in the final handoff.

Ships withagentlas-os

Agent OS: keep specialist agents in a hub, spin up a temporary orchestrator per task. Local-first, works with any model.

Get the whole plugin