메뉴
Reference documentation for the Agent Plugins v1.0.0 specification. Use when you need to understand plugin structure, plugin.json schema, SKILL.md format, mcp.json configuration, or environment variable expansion.
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/kirodotdev/powers --skill agent-plugins-reference명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
Migrate an existing POWER.md-based Kiro Power to the Agent Plugins v1.0.0 specification. Use when converting a power that uses POWER.md frontmatter and steering files to the plugin.json and skills/ format.
Create a new Kiro Power using the Agent Plugins v1.0.0 specification. Use when building a new power from scratch with plugin.json, skills, and optional MCP server configuration.
SOC 직업 분류 기준
SKILL.md 표시 중
| name | agent-plugins-reference |
| description | Reference documentation for the Agent Plugins v1.0.0 specification. Use when you need to understand plugin structure, plugin.json schema, SKILL.md format, mcp.json configuration, or environment variable expansion. |
| license | MIT |
| metadata | {"author":"Kiro Team","version":"1.0.0"} |
Agent Plugins is an open, vendor-neutral specification (v1.0.0) maintained by a Technical Steering Committee including Amazon, Cursor, Microsoft, OpenAI, and Vercel. It defines a portable package format for Agent Skills and MCP servers that compatible clients can discover and load consistently.
A power built to the agent-plugins standard is portable across any conformant client.
my-plugin/
├── plugin.json # Required manifest
├── skills/ # Agent Skills (optional)
│ └── my-skill/
│ ├── SKILL.md # Required per skill
│ ├── scripts/ # Helper scripts (optional)
│ ├── references/ # Reference docs (optional)
│ └── assets/ # Templates, configs (optional)
├── mcp.json # MCP server config (optional)
└── com.example.client/ # Client extensions (optional)
Required. Identifies the plugin and the spec version it targets.
{
"$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json"
Required fields:
$schema - Must be https://agent-plugins.org/schemas/1.0.0/plugin.schema.jsonname - Lowercase, 1-64 chars, only a-z, 0-9, -, .. No consecutive -- or ... Must start/end alphanumeric.Optional metadata fields:
version - Semantic Versioning recommendeddescription - Short descriptionauthor - Object with name, email, url (all optional strings)homepage - Documentation URLrepository - Source repo URLlicense - SPDX identifier recommendedkeywords - Array of search/discovery tagsextensions - Client-specific data keyed by reverse-domain namespaceThe schema defines a fixed set of top-level fields; unknown fields are reported and ignored but do not invalidate the manifest.
Each immediate subdirectory of skills/ containing a SKILL.md file is one skill. Skills follow the Agent Skills specification. Clients do not recursively search deeper descendants.
---
name: "my-skill"
description: "What this skill does. Use when the user needs X."
license: "MIT"
compatibility: "Works with Kiro, Claude Code, Cursor"
metadata:
author: "your-name"
version: "1.0.0"
---
# My Skill
## Overview
What this skill does and when to use it.
## Prerequisites Checklist
- [ ] Requirement 1
- [ ] Requirement 2
## Step-by-Step Guide
### 1. First Step
Instructions with code examples...
## Troubleshooting
### Error: "common error"
**Solution:** How to fix it.
Required frontmatter fields:
name - Lowercase with hyphens, 1-64 chars, must match directory name exactlydescription - Max 1024 chars, start with verb, include when to useOptional frontmatter fields:
license - SPDX identifiercompatibility - Environment requirements (max 500 chars)metadata - Arbitrary key-value pairs (author, version, created, etc.)allowed-tools - Experimental glob patterns for tool restrictionsAll optional:
| Directory | Purpose | Guidelines |
|---|---|---|
scripts/ | Executable helper scripts | Any language, < 500 lines each |
references/ | Additional documentation | API refs, extended examples, checklists |
assets/ | Templates, config files | Static resources referenced from SKILL.md |
Reference files from SKILL.md using relative paths: See [API reference](references/api.md)
| Component | Recommendation | Rationale |
|---|---|---|
| SKILL.md | 400-800 lines | Lightweight discovery and loading |
| Individual scripts | < 500 lines | Easy to understand and modify |
| Total skill directory | < 2 MB | Quick download and indexing |
Move large reference materials to external repositories if needed.
Optional. Configures MCP servers the plugin provides.
{
"$schema": "https://agent-plugins.org/schemas/1.0.0/mcp.schema.json",
"mcpServers": {
"my-server": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@example/mcp-server"],
"env": {
"CONFIG": "${PLUGIN_ROOT}/config.json"
}
}
}
}
Required fields:
$schema - Must be https://agent-plugins.org/schemas/1.0.0/mcp.schema.jsonmcpServers - Object with named server entriesServer transport types:
| Field | Type | Required | Description |
|---|---|---|---|
type | "stdio" | Yes | Selects stdio transport |
command | string | Yes | Single executable token (bare name or ./ relative) |
args | string[] | No | Arguments passed to executable |
env | object of strings | No | Environment variables (cannot contain PLUGIN_ROOT or PLUGIN_DATA keys) |
cwd | string | No | Working directory (must start with ./, ${PLUGIN_ROOT}, or ${PLUGIN_DATA}) |
command must be a single token, not a shell command string./bin/server) resolve against plugin rootcwd is omitted, the plugin root is used| Field | Type | Required | Description |
|---|---|---|---|
type | "streamable-http" | Yes | Selects Streamable HTTP transport |
url | string | Yes | Absolute HTTP/HTTPS URL (HTTPS required for non-loopback) |
headers | object of strings | No | Fixed HTTP headers (no secrets, no expansion) |
| Field | Type | Required | Description |
|---|---|---|---|
type | "sse" | Yes | Selects deprecated HTTP+SSE transport |
url | string | Yes | Absolute HTTP/HTTPS URL |
headers | object of strings | No | Fixed HTTP headers |
mcp.json (bad JSON, unsupported schema version, schema mismatch with plugin.json) disables MCP for that plugin but does not prevent other components from loadingClients provide two variables to stdio subprocesses:
| Variable | Purpose | Contents |
|---|---|---|
${PLUGIN_ROOT} | Bundled read-only files | Absolute path to plugin directory |
${PLUGIN_DATA} | Persistent writable state | Absolute path to client-managed data directory |
Expansion applies to: args elements, env values, cwd string
Expansion does NOT apply to: command, url, header names/values, env keys
Expansion is single-pass, non-recursive. Unrecognized placeholder-like text remains literal. No other placeholder or environment-variable expansion is performed.
Use PLUGIN_ROOT for: referencing bundled scripts, binaries, config files that ship with the plugin.
Use PLUGIN_DATA for: installed dependencies, generated code, caches, persistent state that survives plugin updates.
Client-specific behavior goes in reverse-domain namespaces. Agent Plugins assigns no portable semantics to extension data.
In plugin.json:
{
"$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
"name": "my-plugin",
"extensions": {
"dev.kiro": {
"powerCategory": "development"
}
}
}
As a directory:
my-plugin/
├── plugin.json
├── skills/...
└── dev.kiro/
└── hooks/
Clients ignore extension namespaces they don't implement without validating their contents.
| Kiro Power (old) | Agent Plugin (new) | Notes |
|---|---|---|
| POWER.md frontmatter | plugin.json | name, description, keywords, author move here |
| POWER.md body | skills/*/SKILL.md | Split into discrete skills |
steering/*.md | skills/*/references/*.md or separate skills | Depends on independence |
mcp.json | mcp.json with $schema + type per server | Add schema ref and transport type |
| N/A | skills/*/scripts/ | Executable helper scripts |
| N/A | skills/*/assets/ | Templates and configs |
| N/A | extensions in plugin.json | Client-specific metadata |
| N/A | com.client.name/ directory | Client-specific files |
my-plugin, acme.tools)references/.$schema fieldtype on each server entry${PLUGIN_ROOT} for bundled files, ${PLUGIN_DATA} for generated/cached stateenv or headers