Support for Claude Code plugins and Gemini CLI extensions

[cm-dyoshikawa]

https://www.anthropic.com/news/claude-code-plugins
https://blog.google/technology/developers/gemini-cli-extensions/

When import and generate, can target plugin's (extention's) directory.

[codekiln]

Claude Code Plugins as a Precedent for Rulesync

Just expanding on the idea a bit ...

AI Marketplaces

A repository can act as a marketplace containing multiple plugins, each registering commands, sub-agents, skills, and hooks.
Example marketplaces:

1. ehudhal/claude-code-marketplace
2. anthropics/skills
3. wshobson/agents
4. davila7/claude-code-templates – CLI tool for configuring and monitoring Claude Code
5. GitHub search: **/.claude-plugin/marketplace.json

---

Proposal: Rulesync Marketplaces and Plugins

1. Create .rulesync/marketplace.jsonc

{
  "name": "company-tools",
  "owner": {
    "name": "DevTools Team",
    "email": "devtools@company.com"
  },
  "plugins": [
    {
      "name": "code-formatter",
      "source": "./plugins/formatter",
      "description": "Automatic code formatting on save",
      "version": "2.1.0",
      "author": { "name": "DevTools Team" }
    },
    {
      "name": "deployment-tools",
      "source": {
        "source": "github",
        "repo": "company/deploy-plugin"
      },
      "description": "Deployment automation tools"
    }
  ]
}

Marketplaces can express plugin sources in various ways — see plugin source types:

Relative path

{ "name": "my-plugin", "source": "./plugins/my-plugin" }

GitHub repo

{
  "name": "github-plugin",
  "source": { "source": "github", "repo": "owner/plugin-repo" }
}

---

2. Create Plugin Definition

Add .claude-plugin/plugin.json like claude-code/plugins/agent-sdk-dev/.claude-plugin/plugin.json
(see also the plugins.json spec).
For Rulesync, use .rulesync/plugin.jsonc.

Add supporting folders ./agents/*.md, ./skills/*.md, ./commands/*.md, ./hooks/*.md, and ./.mcp.json.
Based on their sample directory structure:

enterprise-plugin/
├── .rulesync-plugin/           # Metadata directory
│   └── plugin.json             # Required: plugin manifest
├── commands/                   # Default command location
│   ├── status.md
│   └── logs.md
├── subagents/                  # Default agent location
│   ├── security-reviewer.md
│   ├── performance-tester.md
│   └── compliance-checker.md
├── skills/                     # Agent Skills
│   ├── code-reviewer/
│   │   └── SKILL.md
│   └── pdf-processor/
│       ├── SKILL.md
│       └── scripts/
├── hooks/                      # Hook configurations
│   ├── hooks.json
│   └── security-hooks.json
├── .mcp.json                   # MCP server definitions
├── scripts/                    # Hook and utility scripts
│   ├── security-scan.sh
│   ├── format-code.py
│   └── deploy.js
├── LICENSE                     # License file
└── CHANGELOG.md                # Version history

[dyoshikawa]

Claude Code Plugin and Gemini CLI Extension Specification Summary

Claude Code Plugin Specification

Directory Structure:

plugin-name/
├── .claude-plugin/
│   └── plugin.json          # Plugin metadata (required)
├── commands/                 # Slash commands (optional)
│   └── command-name.md
├── agents/                   # Custom agents (optional)
│   └── agent-name.md
├── skills/                   # Agent Skills (optional)
│   └── skill-name/
│       └── SKILL.md
├── hooks/                    # Event handlers (optional)
│   └── hooks.json
├── .mcp.json                 # MCP configuration (optional)
└── README.md                 # Documentation

plugin.json Schema:

{
  "name": "plugin-identifier",
  "description": "Plugin purpose",
  "version": "1.0.0",
  "author": { "name": "Author Name" }
}

Supported Components:

Component	Description
commands	Markdown files in commands/. YAML frontmatter + command content
agents	Markdown files in agents/. Custom agent definitions
skills	skills/<name>/SKILL.md format. Claude uses autonomously based on context
hooks	Event handlers defined in hooks/hooks.json
MCP servers	External tool integration via .mcp.json

Marketplace Distribution (marketplace.json):

{
  "name": "marketplace-name",
  "owner": { "name": "Owner Name" },
  "plugins": [
    {
      "name": "plugin-name",
      "source": "./plugin-path",
      "description": "Plugin description"
    }
  ]
}

References:

• Official Docs
• GitHub
• Blog

---

Gemini CLI Extension Specification

Directory Structure:

<home>/.gemini/extensions/<extension-name>/
├── gemini-extension.json    # Configuration file (required)
├── commands/                # Custom commands (optional)
│   └── command-name.toml
├── GEMINI.md               # Context file (optional)
└── .env                    # Sensitive settings (auto-generated)

gemini-extension.json Schema:

{
  "name": "extension-name",
  "version": "1.0.0",
  "mcpServers": {
    "server-name": {
      "command": "node server.js"
    }
  },
  "contextFileName": "GEMINI.md",
  "excludeTools": ["tool_name"],
  "settings": [
    {
      "name": "Setting Name",
      "description": "Description",
      "envVar": "ENV_VAR_NAME",
      "sensitive": false
    }
  ]
}

Supported Components:

Component	Description
commands	TOML files in commands/. Nested directories create namespaced commands (e.g., commands/gcs/sync.toml → /gcs:sync)
MCP servers	Configured via mcpServers object. All options except trust are supported
context files	Auto-loads contextFileName or GEMINI.md by default
settings	User configuration prompts mapped to environment variables

Configuration Options:

Option	Purpose
name	Unique identifier (lowercase, dashes only)
version	Semantic versioning
excludeTools	Block specific tools or tool parameters
settings	User-provided configuration prompts
contextFileName	Specify custom context filename

Variable Substitution Support:

• ${extensionPath} - Extension path
• ${workspacePath} - Workspace path
• ${/} - Path separator

References:

• Official Docs
• GitHub
• Blog

---

Rulesync Integration Considerations

Feature	Claude Code Plugin	Gemini CLI Extension	Rulesync Mapping
Rules	agents/, skills/	GEMINI.md (context)	.rulesync/rules/ → agents/ or GEMINI.md
Commands	commands/*.md	commands/*.toml	.rulesync/commands/ → Convert to each format
MCP	.mcp.json	Inside gemini-extension.json	.rulesync/mcp.json → Convert to each format
Metadata	plugin.json	gemini-extension.json	New: .rulesync/plugin.json?
Hooks	hooks/hooks.json	Not supported	Existing support is sufficient
Skills	skills/*/SKILL.md	Not supported	Existing support is sufficient

Implementation Approach:

1. rulesync generate --targets plugin - Output Claude Code Plugin format
2. rulesync generate --targets extension - Output Gemini CLI Extension format
3. rulesync import --targets plugin - Import from Plugin
4. rulesync import --targets extension - Import from Extension

[praskr-wisdom]

Has someone started working on this?

[praskr-wisdom]

Has someone started working on this?

[cm-dyoshikawa]

@praskr-wisdom Thank you for your contributing intension. But this issue will be worked by me because many points should be considered.

If you want to contribute this project thankfully, I recommend you to begin to work on good first issues.

[praskr-wisdom]

@dyoshikawa thanks for replying. I asked as I was looking for this feature. I'll take up bugs/features in backlogs relevant to my use case in future. Thanks for building rulesync, it's awesome!

[cm-dyoshikawa]

@praskr-wisdom Really thank you for your positive feedback! I still have strong motivation for this project as long as Rulesync users exist.