Agent Skills Architecture Kit

Up to date (as of 2/27/26) agentic folder protocal and build out guide for OpenClaw, Claude Code, KiloCode, OpenAI Codex, and any AgentSkills-compatible agent. The repo features drag and drop agent folders with 'new-skill' and 'mcp-builder' skills preinstalled to architect any new agent skill or mcp install/creation — ready to use with Claude Code, KiloCode, OpenAI Codex, OpenClaw, and any AgentSkills-compatible agent. 📄 Get the Full Guide [Download the Agent Skills Architecture Guide (PDF)] — covers folder structure, SKILL.md anatomy, frontmatter reference, OpenClaw vs Claude Code comparison diagrams, [claude, kilo, codex, legacy agent] agent folder diagrams, self-hosted model config (Ollama / LM Studio ), troubleshooting, and more.

Follow [@shaneswrld_] on X

---

🚀 Install

Option 1: Download and drag (easiest)

Drop-in .claude/, .kilocode/, .codex/, .agents/, and .agent/ folders with skills, architecture reference, and starter config — ready to use with Claude Code, KiloCode, OpenAI Codex, OpenClaw, and any AgentSkills-compatible agent.

1. Download this repo as a ZIP (green Code button → Download ZIP)
2. Unzip it
3. Drag and drop the folders into your agent folder:
   • For Claude Code: Drag .claude/ into your project root
   • For KiloCode: Drag .kilocode/ into your project root (or use .claude/ as fallback)
   • For OpenAI Codex: Drag .codex/ or .agents/ into your project root
   • For OpenClaw: Drag .agent/ into your project root
   • For AGENT.MD: Drag .agent/AGENT.MD into your project root (or copy to your existing .agent/AGENT.md)
4. Done — your agent now has skills installed and ready to use

How to Drag & Drop MCP Skills

To add MCP skills (like the included mcp-builder skill) to your agent:

1. Download this repo as a ZIP (green Code button → Download ZIP)
2. Unzip it on your computer
3. Navigate to the agent folder you want to add MCP skills to:
   • For OpenClaw: .agent/skills/ folder inside your project
   • For Claude Code: .claude/skills/ folder inside your project
   • For KiloCode: .kilocode/skills/ folder inside your project
   • For OpenAI Codex: .codex/skills/ folder inside your project
4. Drag and drop the skill folder (e.g., new-mcp-builder/ from .agent/skills/new-mcp-builder/) into your agent's skills/ folder
5. Restart your agent — the MCP skill will now be available

Note: The new-mcp-builder skill is located at:

• .agent/skills/new-mcp-builder/ (OpenClaw)
• .claude/skills/new-mcp-builder/ (Claude Code)
• .kilocode/skills/new-mcp-builder/ (KiloCode)

How to Drag & Drop AGENT.MD

To add the AGENT.MD configuration file to your project:

1. Download this repo as a ZIP (green Code button → Download ZIP)
2. Unzip it on your computer
3. Navigate to .agent/ folder in the unzipped repo
4. Drag and drop AGENT.MD into your project's agent folder:
   • For OpenClaw: Drop into your project's .agent/ folder
   • For existing .agent/ folder: Merge with your existing AGENT.md or replace
5. Restart your agent — the agent configuration is now loaded

Tip: If you already have an AGENT.MD file, you can merge the contents or use the included one as a template/reference.

Tip: If you already have an agent folder (.claude/, .kilocode/, .codex/, .agents/, or .agent/), you can just drag the skills/ subfolder into your existing agent folder to add the skills without overwriting your config.

Option 2: Clone

git clone https://github.com/shane9coy/agent-skills-kit.git

# OpenClaw:
cp -r agent-skills-kit/.agent/ your-project/.agent/

# Claude Code:
cp -r agent-skills-kit/.claude/ your-project/.claude/

# KiloCode:
cp -r agent-skills-kit/.kilocode/ your-project/.kilocode/

# OpenAI Codex:
cp -r agent-skills-kit/.codex/ your-project/.codex/

# Legacy Codex / .agents convention:
cp -r agent-skills-kit/.agents/ your-project/.agents/
cp agent-skills-kit/AGENTS.md your-project/AGENTS.md

---

📁 What's Inside

agent-skills-kit/
├── .agent/                                # OpenClaw / general agent
│   ├── AGENT.md                          # Agent configuration and memory
│   └── skills/
│       ├── new-skill-builder/
│       │   ├── SKILL.md                   # Skill installer / scaffolder
│       │   └── references/
│       │       ├── Agent-Skills-Architecture-Guide.md
│       │       └── claude-skills-guide.md
│       └── new-mcp-builder/
│           └── SKILL.md                   # MCP builder skill
│
├── .claude/                              # Claude Code
│   ├── CLAUDE.md                          # Starter project config
│   └── skills/
│       └── new-skill/
│           ├── SKILL.md                   # Skill installer / scaffolder / validator
│           └── references/
│               └── claude-skills-guide.md # Architecture reference for the agent
│
├── .kilocode/                             # KiloCode
│   ├── AGENTS.md                          # KiloCode project memory (AGENTS.md standard)
│   └── skills/
│       └── new-skill/
│           ├── SKILL.md
│           └── references/
│               └── claude-skills-guide.md
│
├── .codex/                                # OpenAI Codex
│   ├── AGENTS.md                          # Codex project memory
│   └── skills/
│       └── new-skill/
│           ├── SKILL.md
│           └── references/
│               └── claude-skills-guide.md
│
├── .agents/                               # Legacy / generic AgentSkills
│   └── skills/
│       └── new-skill/
│           ├── SKILL.md
│           └── references/
│               └── claude-skills-guide.md
│
├── AGENTS.md                              # Codex project memory (root-level fallback)
├── README.md
└── LICENSE

All four folders contain identical skills — use whichever matches your tool. KiloCode also reads .claude/skills/ as fallback, so Claude + KiloCode users only need .claude/. Codex reads .codex/skills/ natively and falls back to .agents/skills/.

---

🛠 What new-skill Does

Once installed, your agent gets 5 workflows:

#	Workflow	How to trigger
1	Create a skill from scratch	"create a new skill for X"
2	Install from GitHub / zip / markdown	"install this skill: <url>"
3	Install the MCP Builder skill	"set up the mcp-builder skill"
4	Validate a skill	"check if my skill is set up correctly"
5	Audit entire agent folder	"scan and fix my .claude/.codex/.kilocode structure"

---

🔀 Compatibility

The AgentSkills spec is a shared standard. This kit works with:

Tool	Agent folder	Skill folder	Memory file
OpenClaw	.agent/	.agent/skills/	AGENT.md
Claude Code	.claude/	.claude/skills/	CLAUDE.md
KiloCode	.kilocode/	.kilocode/skills/ (also reads .claude/skills/)	AGENTS.md (reads CLAUDE.md as fallback)
OpenAI Codex	.codex/	.codex/skills/ (also reads .agents/skills/)	AGENTS.md
OpenCode	.opencode/	.opencode/skills/ (also reads .claude/skills/)	—
Cursor / Windsurf / Cline	.claude/	.claude/skills/ via openskills	—
Any AgentSkills agent	.agents/	.agents/skills/	—

KiloCode also supports mode-specific skills — drop skills into skills-code/, skills-architect/, skills-debug/, etc. to scope them to specific agent modes.

---

📖 Key Concepts (from the PDF guide)

• SKILL.md — the only required file. YAML frontmatter (name + description) for discovery, markdown body for instructions.
• Description is the trigger — the model matches your request against descriptions, not the body. Put all "when to use" info in the description.
• Progressive disclosure — frontmatter is always loaded (~24 tokens/skill), body loads on trigger, references/ load on demand, scripts/ execute on demand.
• Quote your descriptions — unquoted YAML with special characters silently breaks registration.
• Keep SKILL.md under 500 lines — move depth to references/.

---

🤝 Credits

Compiled by s.coy @shaneswrld_ | github.com/shane9coy

---

📜 License

MIT — use it, fork it, ship it.