Post From
Your Editor
Point Claude, Cursor or VS Code at Hibiki and ask for a post. It writes the versions, generates the image and puts them in your queue without you leaving the editor.
What is Hibiki MCP?
The Hibiki MCP server exposes the same tools the dashboard uses, so your AI assistant can draft, format and schedule posts without you switching windows.
Universal Protocol
MCP is an open-source standard that connects AI applications to external systems, like USB-C for AI assistants.
Works Where You Do
Connect Hibiki to Claude Desktop, Claude Code, Cursor, Cline, VS Code, and any MCP-compatible AI tool.
AI-Powered Content
Write and format posts for X, Instagram, LinkedIn, Facebook, Threads, Reddit and your blog.
Complete Workflow
Draft posts, generate images, switch brand styles and schedule the queue, all from the chat window.
How It Works
Install MCP Server
Add Hibiki MCP to your AI tool's configuration
Connect Your API Key
Authenticate with your Hibiki API key
Start Creating
Ask your AI to create content - it handles the rest
Installation
Choose your preferred installation method
Smithery.ai
Easiest: One-click installation with automatic updates
Open Smithery.aiSteps:
- 1.Visit Smithery.ai page
- 2.Click "Install" for your AI tool
- 3.Provide your Hibiki API key when prompted
- 4.Configuration applied automatically
NPX
Run directly with npx without installation
npx @hibiki/mcp-server --api-key YOUR_API_KEYGet Your API Key
You'll need a Hibiki API key to use the MCP server. Get yours from the developer dashboard.
Get API KeySee It In Action
Pick a use case and watch the exchange: the tool call that went out, and the drafts that came back
Claude with Hibiki MCP
Content Creator Automation Demo
💡 Click any use case above to see a different demo
Set It Up In Two Minutes
Add the MCP server to Claude, Cursor or VS Code, paste your API key, and ask for a post. The free plan is enough to try the whole loop.
Setup Guides
Configure Hibiki MCP for your preferred AI tool
Claude Desktop
Use Hibiki with Claude Desktop app
~/Library/Application Support/Claude/claude_desktop_config.json{
"mcpServers": {
"hibiki": {
"command": "npx",
"args": [
"@hibiki/mcp-server",
"--api-key",
"YOUR_HIBIKI_API_KEY"
]
}
}
}Configuration Steps:
- 1Open Claude Desktop settings
- 2Navigate to Developer → Edit Config
- 3Add the configuration shown above
- 4Replace YOUR_HIBIKI_API_KEY with your actual key
- 5Save and restart Claude Desktop
Claude Code
Integrate Hibiki with Claude Code CLI
~/.config/claude/config.json{
"mcpServers": {
"hibiki": {
"command": "npx",
"args": [
"@hibiki/mcp-server",
"--api-key",
"YOUR_HIBIKI_API_KEY"
]
}
}
}Configuration Steps:
- 1Locate your Claude Code config file
- 2Add the MCP server configuration
- 3Insert your Hibiki API key
- 4Restart Claude Code
VS Code
Use Hibiki with VS Code MCP extension
VS Code Settings → MCP Configuration{
"mcpServers": {
"hibiki": {
"command": "npx",
"args": [
"@hibiki/mcp-server",
"--api-key",
"YOUR_HIBIKI_API_KEY"
]
}
}
}Configuration Steps:
- 1Install MCP extension from VS Code marketplace
- 2Open VS Code settings (Cmd/Ctrl + ,)
- 3Search for "MCP"
- 4Add MCP server configuration
- 5Save and reload window
OpenAI Codex
Integrate Hibiki with OpenAI Codex CLI and IDE extension
Install via Smithery~/.codex/config.toml[mcp.hibiki]
command = "npx"
args = [
"@hibiki/mcp-server",
"--api-key",
"YOUR_HIBIKI_API_KEY"
]Configuration Steps:
- 1Install OpenAI Codex CLI
- 2Run "codex mcp add hibiki -- npx @hibiki/mcp-server --api-key YOUR_API_KEY"
- 3Or manually edit ~/.codex/config.toml
- 4Replace YOUR_HIBIKI_API_KEY with your actual key
- 5Configuration syncs between CLI and IDE extension
Cline
Set up Hibiki in Cline VS Code extension
Install via SmitheryVS Code Settings → Cline → MCP Settings{
"mcpServers": {
"hibiki": {
"command": "npx",
"args": [
"@hibiki/mcp-server",
"--api-key",
"YOUR_HIBIKI_API_KEY"
]
}
}
}Configuration Steps:
- 1Open VS Code settings (Cmd/Ctrl + ,)
- 2Search for "Cline MCP"
- 3Click "Edit in settings.json"
- 4Add the configuration
- 5Reload VS Code window
Cursor
Enable Hibiki in Cursor AI IDE
Cursor Settings → Features → MCP{
"mcpServers": {
"hibiki": {
"command": "npx",
"args": [
"@hibiki/mcp-server",
"--api-key",
"YOUR_HIBIKI_API_KEY"
]
}
}
}Configuration Steps:
- 1Open Cursor settings
- 2Navigate to Features → MCP
- 3Add new MCP server
- 4Paste configuration and save
- 5Restart Cursor
ChatGPT Desktop
Use Hibiki with ChatGPT Desktop app (Developer Mode)
Install via SmitheryChatGPT Settings → Developer Mode{
"mcpServers": {
"hibiki": {
"command": "npx",
"args": [
"@hibiki/mcp-server",
"--api-key",
"YOUR_HIBIKI_API_KEY"
]
}
}
}Configuration Steps:
- 1Open ChatGPT Desktop app
- 2Go to Settings → Developer Mode (Beta)
- 3Enable Developer Mode
- 4Add MCP server configuration
- 5Restart ChatGPT Desktop
Raycast
Integrate Hibiki with Raycast AI
Raycast → Manage MCP Servers{
"name": "Hibiki",
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@hibiki/mcp-server",
"--api-key",
"YOUR_HIBIKI_API_KEY"
]
}Configuration Steps:
- 1Open Raycast and search for "Install Server"
- 2Fill in the server configuration form
- 3Or use deep link for one-click install
- 4@-mention Hibiki in AI Chat to use
- 5View config: "Manage MCP Servers" → "Show Config File"
Gemini CLI
Use Hibiki with Google Gemini CLI
Install via Smithery~/.gemini/settings.json{
"mcpServers": {
"hibiki": {
"command": "npx",
"args": [
"@hibiki/mcp-server",
"--api-key",
"YOUR_HIBIKI_API_KEY"
]
}
}
}Configuration Steps:
- 1Install Gemini CLI: npm install -g @google/gemini-cli@latest
- 2Edit ~/.gemini/settings.json
- 3Add Hibiki MCP server configuration
- 4Replace YOUR_HIBIKI_API_KEY with your actual key
- 5Restart Gemini CLI
Goose
Integrate Hibiki with Goose AI agent
Install via SmitheryInteractive: goose configure# Run: goose configure
# Choose: Add Extension → Remote Extension
# Name: Hibiki
# SSE Endpoint: <from Smithery>
# Timeout: 300Configuration Steps:
- 1Install Goose CLI
- 2Run "goose configure"
- 3Select "Add Extension" → "Remote Extension"
- 4Name it "Hibiki"
- 5Use Smithery for SSE endpoint URL
💡 Pro Tip
After configuring, test the connection by asking your AI assistant: "Can you create a post about AI using Hibiki?" Your assistant should be able to use the Hibiki MCP tools to generate content.