Agents Teamwork — Development skill for Claude Code
Universal multi-IDE agent teamwork skills: one team-orchestrator skill + per-tool role configs for 12 coding tools (Claude, Codex, Copilot, Cursor, Windsurf, Trae, Qoder, Antigravity…).
How to install Agents Teamwork
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open jessexu1818/agents-teamwork and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Agents Teamwork does
Universal multi-IDE agent teamwork skills: one team-orchestrator skill + per-tool role configs for 12 coding tools (Claude, Codex, Copilot, Cursor, Windsurf, Trae, Qoder, Antigravity…).
Alternatives in Development
- Agentic Stack — One brain, many harnesses 2.2k ★
- Chops — 1,000+ macOS app to browse, edit, and manage skills across Claude Code, Cursor, Codex, Windsurf, and Amp 1.1k ★
- HERO Anti OverDefense — HERO = Hashing · Edge cases · Rubrics · Overbuild — the four shapes coding agents over-defend in 333 ★
README
agents-teamwork
[]() []() []() []()

Universal multi-IDE agent teamwork skills: one `team-orchestrator` skill plus per-tool role configs that turn any supported IDE into a root-plus-specialists team. The root owns architecture, decomposition, integration, verification, and the final answer; subagents supply bounded evidence and execution only.
Why
- One skill file works everywhere, even with zero generated config (skills-only mode).
- Generated role files prefill model, sandbox, and routing defaults per tool.
- Explicit delegation contracts keep parallel subagents from stepping on each other: one writer per file or subsystem, root mediates ownership.
- Category routing (
quick/deep/ultrabrain/visual) scales cost to task difficulty.
Topology
team-orchestrator (root)
architect + integrator
____________________|____________________
| | | | |
explorer worker tester reviewer researcher (core)
| | | |
planner oracle designer ........ (optional extras)
(.plans/ counsel UI/vision
interview)
The root delegates through bounded contracts (objective, scope, context, constraints, deliverable, acceptance) and merges the returned slices. Never claim delegation without a real subagent call.
Quickstart
Three options, in order. Start with option 1.
(1) Skills-only install — FIRST, 30 seconds, zero scripts
Copy the single skill file into your repo. That file alone is enough: the root can play e
Related Skills
Coding2API
把 CodeBuddy(国内站 + 国际版)、TRAE SOLO、OpenCode Zen、Kilo Gateway、Qoder、CodeArts、OfficeAce 八个上游渠道封装为 OpenAI / Respons
Face
FACE, Focus Agentic Code Environment: the window a team opens when coding agents write the code. One free bina
HiveForge
Autonomous Multi-CLI Agent Team Mission Control & Orchestrator. Coordinate Google Antigravity, Claude Code, Op
Trae Superpowers
Install obra/superpowers skills into the Trae IDE, activated from the first message via User Rules.
Jgbriel Skills
IDE configs, skills, and steering for Claude Code + Kiro IDE.
Setup Stack
Generates stack-specific agents, skills, and configs, creating the active stack (config.json) without touching
Related Agents
Agentic Auditor
PYA specialist spawned by the /pya orchestrator. Audits AI agent and LLM surfaces in an untrusted repository:
Contract Auditor
Reads one teamwork team root with fresh context and reports where a contract has stopped being a contract — a
Developer Overview
code-container (container) creates isolated Docker environments for AI coding harnesses (Claude Code, OpenCode