By JsonLee12138
Orchestrate multi-agent AI teams for collaborative software development workflows: bootstrap projects, create and manage roles, decompose tasks, dispatch workers, inspect status, ensure QA processes, and recover sessions via CLI commands.
Compatibility shell for the legacy umbrella agent-team skill. Use only when older prompts still reference `agent-team`; route the request to the dedicated scenario skill instead of treating this file as the primary execution surface.
Use when users explicitly ask to brainstorm, shape requirements, compare approaches, or produce a planning/design document before implementation. Turn rough ideas into validated brainstorming/design docs through focused dialogue, role-based analysis, and explicit user approval. Do not trigger this skill for straightforward implementation requests that do not need dedicated design exploration.
Read-only catalog browsing skill for searching and viewing role catalog information. Use when the user wants to list, search, inspect, or review catalog repositories and statistics.
Session context cleanup and file re-anchoring strategy for controller and worker sessions. Use when the session is drifting, phases change, or resumed work needs index-first recovery from files.
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
English | 中文
Orchestrate your AI workforce with surgical precision. 🚀
agent-team is a multi-agent development manager that uses a Role + Worker model to run AI agents in isolated Git worktrees and dedicated terminal sessions.
.agent-team/teams/) defining goals, prompts, and tools..worktrees/) with their own branch and session.Let your AI agent set itself up. It only takes two steps:
npx skills add JsonLee12138/agent-team -a <platform> -y
Replace <platform> with your provider: claude, gemini, opencode, or codex."Install agent-team and initialize the project."
| Method | Command |
|---|---|
| Homebrew | brew tap JsonLee12138/agent-team && brew install agent-team |
| Go Install | go install github.com/JsonLee12138/agent-team@latest |
| Claude Plugin | /plugin marketplace add JsonLee12138/agent-team |
| Gemini Ext | gemini extensions install https://github.com/JsonLee12138/agent-team |
| OpenCode Plugin | { "plugin": ["opencode-agent-team"] } in ~/.config/opencode/opencode.json |
| Method | Command |
|---|---|
| Claude Plugin | /plugin marketplace update agent-team |
| Skill | npx skills add JsonLee12138/agent-team -a '*' -y |
| Homebrew | brew update && brew upgrade agent-team |
| Go Install | go install github.com/JsonLee12138/agent-team@latest |
Manage your entire team through natural language. Your AI agent will handle the commands for you.
"Create a frontend-architect role to manage our UI infrastructure."
Pull in expert roles from this repository:
agent-team role-repo add JsonLee12138/agent-team
"Create a worker for frontend-architect using Claude." This opens a new terminal window with an isolated worktree.
"Assign frontend-architect-001 to design the new auth flow." The agent will brainstorm a design doc before touching any code.
"Merge frontend-architect-001 and delete the worker."
Available in .agent-team/teams/:
pm: Product Manager & Requirement Shaping.frontend-architect: High-level UI/UX structure.vite-react-dev: Specialized for Vite + React.pencil-designer: UI design tool specialist.task-orchestrator: Task lifecycle entry.workflow-orchestrator: Governance-only workflow plan entry.worker-dispatch: Controller-side worker open/reply entry.worker-recovery: Worker-first resume entry.worker-reply-main: Worker -> main reporting entry.context-cleanup: Session cleanup + index-first file re-anchoring.task-inspector: Read-only task inspection.role-repo-manager: Role source management.catalog-browser: Read-only catalog browsing.project-bootstrap: init / migrate entry.rules-maintenance: rules sync entry.skill-maintenance: Skill cache maintenance.worker-inspector: Read-only worker status.role-browser: Read-only local role browsing.role-creator: Interactively build new agent roles.brainstorming: Validates ideas via dialogue before implementation, including planning-target selection (roadmap / milestone / phase / task / generic topic) and explicit save-destination choices (docs/brainstorming/, target-object directory, custom path, or skip saving).agent-team: Legacy compatibility shell that now routes to the dedicated scenario skills.| Provider | CLI Value | Integration |
|---|---|---|
| Claude Code | claude | Plugin |
| Gemini CLI | gemini | Extension |
| OpenCode | opencode | NPM Plugin |
| OpenAI Codex | codex | Prompt-only |
npx claudepluginhub jsonlee12138/agent-team --plugin agent-teamMulti-agent team orchestration for Claude Code. Set up parallel AI agent teams with file-based planning, progress tracking, and role-based collaboration.
Intelligently compose and deploy Claude Code Agent Teams. Auto-selects optimal team composition from 24+ agents and 83+ skills across 5 scopes, generates task dependency graphs, and orchestrates multi-agent workflows with a single command.
Orchestrate multi-agent teams for parallel code review, hypothesis-driven debugging, and coordinated feature development using Claude Code's Agent Teams
Multi-agent team orchestration for parallel task execution, research, and implementation
AI Agent Team Operating System for Claude Code — persistent team management, meetings, task wall, company loop engine, and real-time dashboard
Multi-agent orchestration for Claude Code. 12 specialized agents working in parallel — planning, building, reviewing, debugging. Plus a Hub for always-alive multi-project sessions controllable from Telegram or Slack.