Use Docker-aware agents for containerized development:
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 →Use Docker-aware agents for containerized development:
Define the specialized agents your AI assistant can delegate to. Each agent has a focused purpose and clear trigger conditions.
> **IMPORTANT**: All code in this project MUST follow these principles.
`agents/` directory at plugin root
When working with Kailash frameworks, MUST consult the relevant specialist:
Located in `.claude/agents/` (project-level):
Claudekit includes 32 specialized AI subagents that provide deep domain expertise. These agents work alongside Claude Code to handle complex, domain-specific tasks.
This file declares the available sub-agents (skills) for the Operational Excellence (OpEx) BPM agent.
These rules apply to all roles. Role files may add detail but cannot override them.
Scans your Next.js codebase for patterns that drive up your Vercel bill, focusing on compute duration, function invocations, and bandwidth optimization.
This is a **production-ready AI coding setup** providing 13 specialized agents, 30 skills, 21 commands, and automated hook workflows for software development.
<!-- Parent: ../AGENTS.md -->
These agents have been updated to integrate with your boilerplate system's architecture, including:
> CLAUDE.mdから分離。各サブエージェントの役割と発火タイミングを定義。
> **Level: 🌳 Expert** | **Source:** [Sub Agents](https://docs.anthropic.com/en/docs/claude-code/sub-agents)
Generate project-specific AGENTS.md files with custom context derived from project analysis or user description.
How to set up and use Claude Code custom agents for game development — specialized sub-agents for engineering, narrative, and design work.
- **Time-based scheduling** — Schedule tasks daily, weekly, or at specific times
Temp agents are ephemeral, workflow-specific agents that I create automatically when you need specialized functionality.
Descrição dos agentes especializados para análise de PRs.
Subagents are specialized AI instances that Claude delegates tasks to. Each runs in its own context window with custom system prompt, specific tool access, and independent permissions.
<!-- This file is an INDEX, not an encyclopedia. Keep it under 150 lines.
General instructions for AI coding agents working on this project.
Sources: Huyen (AI Engineering, ch. 6 & 10), Lanham (AI Agents in Action), Arsanjani & Bustos (Agentic Architectural Patterns), 2025–2026 production patterns
Expert systems and intelligent agents face a fundamental paradox: the knowledge that makes expert performance *excellent* is precisely the knowledge that is hardest to specify, capture, and verify. Th
{{PROJECT_NAME}} is managed with a brief-first, Specify-first workflow pack.
All agents are spawned as `subagent_type: "general-purpose"`. All agents are **read-only** — they return structured markdown reports, no edits.
Working with subagents in the Claude Agent SDK
> ## Documentation Index
The `SequentialAgent` is a workflow agent that executes its sub-agents in the order they are specified in the list.
Quick reference for programmatic usage.
Bond is a cross-platform, schema-first serialization framework developed by Microsoft. It defines data schemas in `.bond` files that are compiled into language-specific classes, supporting C#, C++, Py
Build buyer and seller agent workflows with Skyfire KYA, PAY, and KYA+PAY tokens. Use when implementing token creation, token introspection and charging, seller service lifecycle, service discovery, Skyfire MCP integration, or enterprise admin operations.
**Version:** 1.0.0
> **Note:**
> Comprehensive guide for authoring `.swirls` workflow files. Compiled from the individual rule files under `rules/`. If something is not documented here, it does not exist in the language.
Fake JSON Server provides lightweight, in-process HTTP servers for testing .NET applications that depend on external REST APIs. Instead of calling real third-party services during tests, you create fa
Python Django + React TypeScript application.
QA 自动化工程师,精通 Playwright/Puppeteer,擅长编写稳定、可维护的端到端测试。
Complete guide for building Lobster AI agents — from scoping through PR.
You are GEO-Claw 🔍, an AI visibility optimization specialist. Your skill is loaded at:
Lobster AI uses specialist agents for different analysis domains. The Supervisor
Activate this skill when:
(Use the version from your main starter repo. This file defines Architect, Builder, Tester roles, routing rules, workflow, coding standards, and done criteria.)
Agents are specialized roles your AI system plays. Each agent has a defined persona, explicit tool permissions, a list of skills it can execute, and hard boundaries on what it must NOT do.
Subagents in the Claude Agent SDK are specialized AIs that are orchestrated by the main agent. Use subagents for context management and parallelization.
Software specifications capture the "what" and "why" before implementation addresses the "how." Well-structured requirements documents, architecture decisions, and executable specs reduce ambiguity, a
A Claude Code skill that enables efficient reading of EPUB ebook files.
AutoFixture is a .NET library that minimizes the "Arrange" phase of unit tests by automatically generating test data for any type. It creates anonymous instances of classes, structs, primitives, colle
Control plane for Obsidian vault operations. Manages skill-based routing to specialized reference files. Reduces context overhead by loading only relevant guidance.