Spawn subagents to develop features in parallel worktrees
Skills（SKILL.md）は、AIエージェント（Claude Code、Cursor、Codexなど）に特定の能力を追加するための設定ファイルです。
詳しく見る →Spawn subagents to develop features in parallel worktrees
- **Providers**: register in `src/models/registry.ts#initProviders`. Add the
An agent in Microsoft Agent Framework is a stateful conversational entity that combines:
Instructions for any AI coding agent working on this project.
 .
> **Scope:** AI coding agents (Claude, Cursor, Copilot, and any other AI assistant operating in this codebase).
자율 AI 코딩 에이전트 저장소 목록 (Stars 순)
- **Independent searches**: Looking for multiple unrelated things in a codebase
- [Creating Agents](#creating-agents)
Always set `model:` when spawning subagents. Haiku costs 12× less than Sonnet — use it for all search and explore work.
This file documents all 5 Agno-powered agents in the WAIGenie framework.
Create AI agents with custom instructions, tools, and model configurations.
<!-- Repo-level context for AI coding agents (Gemini, Claude, open-swe, etc.)
The vocabulary you didn't know you needed. 1 skill, 17 commands, and curated anti-patterns for impeccable style. Works with Cursor, Claude Code, Gemini CLI, and Codex CLI.
Bu, yazılım geliştirme için 28 özel agent, 116 skill, 59 command ve otomatik hook iş akışları sağlayan **üretime hazır bir AI kodlama eklentisidir**.
This file defines the lead-agent operating model for Aegis feature work in Claude-driven sessions.
How to use different LLM models for different agents in Present-Agent2.
このスキルでは、Task toolを使用してgeneral-purpose subagentにプロンプトを渡す形式で処理を行う。
| Agent ID | Owns | Entry dirs | Model | Can parallel with |
We investigate the challenge of task planning for multi-task embodied agents in open-world environments.[^2] Two main difficulties are identified: 1) executing plans in an open-world environment (e.g.
| Agent | subagent_type | isolation | run_in_background |
This command provides an overview of all available development agents and their status.
> This file defines the agent team structure, sub-agent roles, and orchestration patterns for the MTA Website project.
This directory is the canonical home for all agent definitions in this repo. Both Claude Code and OpenCode reference these files as the single source of truth.
이 문서는 AI 가계부 프로젝트의 모든 코딩 컨벤션을 정의합니다.
这是一个**生产就绪的 AI 编码插件**，提供 28 个专业代理、116 项技能、59 条命令以及自动化钩子工作流，用于软件开发。
This repository uses **layered documentation** for AI agent context management.
This project uses Claude Code project settings (in `.claude/settings.json`) to enforce constraints deterministically.
<!--
List all defined agents with their roles, tiers, and teams
Thaler and Sunstein's choice architecture maps directly to hook enforcement design -- blocking hooks are mandates, context-injecting hooks are nudges, and the graduation between them prevents alert
> Operating instructions for OpenClaw agents working with Claude Code Autopilot projects.
Markdown + YAML Frontmatter形式で記述します。
> 来源：CLAUDE.md v2.0
<!-- FOR AI AGENTS - Human readability is a side effect, not a goal -->
> **FOR AI AGENTS**: This document provides coding standards and best practices for TypeScript/JavaScript development. Follow these conventions for all code contributions.
Claude Code internal subagents. These are NOT end-user agents — they're
Automatically spawn the right agents at the right time without manual intervention.
This reference is shared across all agents. Every agent knows the others, their responsibilities, and when to suggest them to the dispatcher.
Sub-agents let you run background tasks without blocking the main conversation. When you spawn a sub-agent, it runs in its own isolated session, does its work, and announces the result back to the cha
Agent configurations for Claude CLI. Tools use Claude's tool names.
Advanced pattern for Ralph Loop using sub-agents, multiple models, and adversarial verification.
**Repo:** https://github.com/wshobson/agents
> ## Documentation Index
This reference defines 5 coaching agent personas used by the `/octave:deal-coach` skill. Three primary agents map to the Resonate → Elevate → Compel stages. Two cross-stage agents provide supplemental
- [Agent CRUD](#agent-crud)
This is the highest-density area of unknown unknowns because AI agent systems are new enough that most developers have never encountered these failure modes.
Use this package when the user wants persuasive marketing copy such as ads, landing page copy, launch messaging, headlines, or CTA variants.
Parallel task execution via Agent tool dispatches with worktree isolation. No experimental env var needed.