Sooshie

Sooshie MCP Server

An MCP (Model Context Protocol) server that lets AI assistants like Claude manage your social media through Sooshie — schedule posts, generate AI captions, manage accounts, and more.

What is this?

Connect your AI assistant to your Sooshie workspace and manage social media through natural conversation:

• "Schedule a post about our new product launch for tomorrow at 9am on Instagram and LinkedIn"
• "What posts do I have scheduled for this week?"
• "Generate a witty caption about our summer sale for TikTok"
• "How many failed posts do I have?"

Tools (29)

Posts

Tool	Description
list_posts	List posts with optional status/date filters
get_post	Get a specific post by ID
create_post	Create a new post (draft, scheduled, or publish now)
update_post	Update post content, media, or targets
schedule_post	Schedule a post for a specific date/time
publish_post	Publish a post immediately

Tool	Description
get_post_comments	Get comments on a post
add_post_comment	Add a comment to a post
delete_post_comment	Delete a comment
submit_post_for_review	Submit a post for approval
review_post	Approve or request changes on a post

Bulk Operations

Tool	Description
delete_posts	Bulk delete multiple posts
bulk_import_csv	Import posts from a CSV file
bulk_reschedule_posts	Reschedule multiple posts at once

Post Logs

Tool	Description
get_post_logs	Get activity log for a post
get_failed_post_count	Get count of failed posts

Accounts

Tool	Description
list_accounts	List connected social accounts
check_account_health	Check connection status of an account
delete_account	Disconnect a social account

Workspaces

Tool	Description
list_workspaces	List all workspaces
get_workspace	Get current workspace details
update_workspace	Update workspace settings
get_workspace_members	List workspace team members
invite_member	Invite a team member by email

AI Content

Tool	Description
generate_content	Generate a social media caption with AI
rewrite_content	Rewrite existing content in a different tone
get_best_times	Get recommended posting times per platform

Media & API Keys

Tool	Description
upload_media	Upload an image or video
delete_media	Delete uploaded media
list_api_keys	List API keys
revoke_api_key	Revoke an API key

Supported Platforms

Instagram, Facebook, LinkedIn, TikTok, YouTube, Threads, X (Twitter), Pinterest, Reddit, Bluesky, Google Business Profile, Telegram, Snapchat

Setup

1. Get your API key

1. Sign up at sooshie.app
2. Go to Settings > API Keys
3. Create a new API key (starts with sk_live_)

2. Install

npm install -g @sooshie/mcp-server

Or run directly with npx:

npx @sooshie/mcp-server

3. Configure Claude Desktop

Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "sooshie": {
      "command": "npx",
      "args": ["-y", "@sooshie/mcp-server"],
      "env": {
        "SOOSHIE_API_KEY": "sk_live_your_key_here",
        "SOOSHIE_URL": "https://sooshie.app"
      }
    }
  }
}

4. Restart Claude Desktop

The Sooshie tools will appear in Claude's tool list.

Environment Variables

Variable	Required	Default	Description
SOOSHIE_API_KEY	Yes	—	Your Sooshie API key (sk_live_...)
SOOSHIE_URL	No	https://sooshie.app	Sooshie instance URL

Examples

Create a post:

"Create a draft post about our upcoming webinar on social media strategy. Target Instagram and LinkedIn."

Schedule content:

"Schedule a post for next Tuesday at 10am EST: 'Excited to announce our new AI features! Try them free at sooshie.app'"

Generate captions:

"Generate 3 different captions for a photo of our team volunteer event. Make them professional for LinkedIn."

Check analytics:

"What posts did I publish this week? Which ones are scheduled for next week?"

Manage team:

"Invite sarah@company.com as an admin to our workspace"

Development

# Install dependencies
npm install

# Build
npm run build

# Run tests
npm test

# Watch mode
npm run dev

License

MIT

Links

• Sooshie — Social media scheduling platform
• Documentation — Full setup guide
• MCP Protocol — Model Context Protocol specification