Plugin Creator
About
This Claude Skill automatically scaffolds new Claude Code plugins with proper structure and marketplace integration when triggered by phrases like "create plugin" or "plugin from template." It handles the complete setup process including directory creation, file generation, and catalog updates specifically for the claude-code-plugins repository workflow. Developers should use it to quickly generate validated plugin templates with all required components.
Documentation
Plugin Creator
Purpose
Automatically scaffolds new Claude Code plugins with complete directory structure, required files, proper formatting, and marketplace catalog integration - specifically optimized for the claude-code-plugins repository.
Trigger Keywords
- "create plugin" or "new plugin"
- "plugin from template"
- "scaffold plugin"
- "generate plugin"
- "add new plugin to marketplace"
Plugin Creation Process
When activated, I will:
-
Gather Requirements
- Plugin name (kebab-case)
- Category (productivity, security, devops, etc.)
- Type (commands, agents, skills, MCP, or combination)
- Description and keywords
- Author information
-
Create Directory Structure
plugins/[category]/[plugin-name]/ ├── .claude-plugin/ │ └── plugin.json ├── README.md ├── LICENSE └── [commands|agents|skills|hooks|mcp]/ -
Generate Required Files
- plugin.json with proper schema (name, version, description, author)
- README.md with comprehensive documentation
- LICENSE (MIT by default)
- Component files based on type
-
Add to Marketplace Catalog
- Update
.claude-plugin/marketplace.extended.json - Run
npm run sync-marketplaceautomatically - Validate catalog schema
- Update
-
Validate Everything
- Run
./scripts/validate-all.shon new plugin - Check JSON syntax with
jq - Verify frontmatter in markdown files
- Ensure scripts are executable
- Run
Plugin Types Supported
Commands Plugin
- Creates
commands/directory - Generates example command with proper frontmatter
- Includes
/demo-commandexample
Agents Plugin
- Creates
agents/directory - Generates example agent with capabilities
- Includes model specification
Skills Plugin
- Creates
skills/skill-name/directory - Generates SKILL.md with proper format
- Includes trigger keywords and allowed-tools
MCP Plugin
- Creates
src/,dist/,mcp/directories - Generates TypeScript boilerplate
- Includes package.json with MCP SDK
- Adds to pnpm workspace
Full Plugin
- Combines all types
- Creates complete example structure
- Ready for customization
File Templates
plugin.json Template
{
"name": "plugin-name",
"version": "1.0.0",
"description": "Clear description",
"author": {
"name": "Author Name",
"email": "[email protected]"
},
"repository": "https://github.com/jeremylongshore/claude-code-plugins",
"license": "MIT",
"keywords": ["keyword1", "keyword2"]
}
Command Template
---
name: command-name
description: What this command does
model: sonnet
---
# Command Title
Instructions for Claude...
Skill Template
---
name: Skill Name
description: What it does AND when to use it
allowed-tools: Read, Write, Grep
---
# Skill Name
## Purpose
[What this skill does]
## Trigger Keywords
- keyword1
- keyword2
## Instructions
[Step-by-step for Claude]
Marketplace Integration
I automatically:
- Add plugin entry to
marketplace.extended.json - Run
npm run sync-marketplaceto update CLI catalog - Validate both catalogs with
jq - Check for duplicate names
- Verify source paths exist
Validation Steps
After creation:
- ✅ All required files present
- ✅ Valid JSON (plugin.json, catalogs)
- ✅ Proper frontmatter in markdown
- ✅ Scripts executable (
chmod +x) - ✅ No duplicate plugin names
- ✅ Category is valid
- ✅ Keywords present
Repository-Specific Features
For claude-code-plugins repo:
- Follows exact directory structure
- Uses correct marketplace slug (
claude-code-plugins-plus) - Includes proper LICENSE file
- Adds to correct category folder
- Validates against existing plugins
- Updates version in marketplace
Output
I provide:
✅ Created plugin: plugin-name
📁 Location: plugins/category/plugin-name/
📝 Files created: 8
🔍 Validation: PASSED
📦 Marketplace: UPDATED
✨ Ready to commit!
Next steps:
1. Review files in plugins/category/plugin-name/
2. Customize README.md and component files
3. Run: git add plugins/category/plugin-name/
4. Run: git commit -m "feat: Add plugin-name plugin"
Examples
User says: "Create a new security plugin called 'owasp-scanner' with commands"
I automatically:
- Create directory:
plugins/security/owasp-scanner/ - Generate plugin.json, README, LICENSE
- Create
commands/with example - Add to marketplace.extended.json
- Sync marketplace.json
- Validate all files
- Report success
User says: "Scaffold a Skills plugin for code review"
I automatically:
- Create directory with
skills/subdirectories - Generate SKILL.md templates
- Add trigger keywords for code review
- Add to marketplace
- Validate and report
Quick Install
/plugin add https://github.com/jeremylongshore/claude-code-plugins-plus-skills/tree/main/plugin-creatorCopy and paste this command in Claude Code to install this skill
GitHub 仓库
Related Skills
sglang
MetaSGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.
evaluating-llms-harness
TestingThis Claude Skill runs the lm-evaluation-harness to benchmark LLMs across 60+ standardized academic tasks like MMLU and GSM8K. It's designed for developers to compare model quality, track training progress, or report academic results. The tool supports various backends including HuggingFace and vLLM models.
llamaguard
OtherLlamaGuard is Meta's 7-8B parameter model for moderating LLM inputs and outputs across six safety categories like violence and hate speech. It offers 94-95% accuracy and can be deployed using vLLM, Hugging Face, or Amazon SageMaker. Use this skill to easily integrate content filtering and safety guardrails into your AI applications.
langchain
MetaLangChain is a framework for building LLM applications using agents, chains, and RAG pipelines. It supports multiple LLM providers, offers 500+ integrations, and includes features like tool calling and memory management. Use it for rapid prototyping and deploying production systems like chatbots, autonomous agents, and question-answering services.
