Luftsocial API
Two ways to automate Luftsocial: REST API for backend integrations and scripts, and an MCP Server that AI agents like Cursor and Claude call directly using natural language. Both use the same API key.
Open MCP settings
Create API keys, copy MCP config, manage connected clients
Start Building with AI
Prompt readyPaste into your AI assistant to read our docs and start coding.
You are helping me build an integration with the Luftsocial API. Please start by reading the developer documentation at: https://luftsocial.com/developers Key sections to read: - Overview & Quick Start - REST API Reference - MCP Server Setup (for AI agent integrations) - Authentication & API Keys After reading the docs, help me: 1. Understand the available API endpoints and MCP tools 2. Set up authentication with my API key 3. Build the integration I describe below My integration goal: [describe what you want to build]
Quick Start
Four steps from zero to your first published post.
Create an API Key
Dashboard → Settings → Developer → New API Key. Choose a scope preset and copy the value immediately — shown once.
Connect an AI Agent via MCP
MCP ServerPaste your API key into Cursor, Claude Desktop, Codex, or Windsurf. 33 tools callable in plain English.
https://mcp.luftsocial.com/mcpHow it fits together
Same key, two surfaces — the prefix routes traffic automatically.
Integration Architecture
/api/v1/agent/…← REST API
33 tools across 4 groups← MCP Server
Both surfaces authenticate with Authorization: Bearer ls_live_… — the key prefix routes to production or staging automatically.
REST API vs MCP
Use REST when writing HTTP code directly — scripts, automation pipelines, backend services. Use MCP when working inside an AI assistant and want it to post and schedule for you without writing any HTTP code.
| REST API | MCP Server | |
|---|---|---|
| Best for | Backend scripts, automation pipelines, n8n workflows, custom server integrations | AI coding assistants — Cursor, Claude Desktop, Codex, Windsurf |
| How you interact | HTTP requests to typed JSON endpoints (curl, fetch, requests) | Natural-language prompts — the AI calls the right tool automatically |
| Authentication | Authorization: Bearer ls_live_… — one header, no extras needed | Same API key in an Authorization header — or OAuth (paste URL only) |
| Docs & tooling | OpenAPI reference (Scalar) — browse & test endpoints interactively | MCP setup wizard + searchable tool catalog with 33 tools |
| Setup time | ~2 min — create a key, export env var, make your first request | ~3 min — create a key, paste config snippet into your AI client, ask a question |
| Subscription | Pro or higher required | Pro or higher required |
Quick start steps
Follow in order for the fastest path to a working integration.
- 1
Read the docs (no account required)
Browse the API reference and MCP tool catalog without signing up. Sign in to a workspace when ready to make real requests. MCP uses your plan.
- 2
Understand your key prefix
ls_live_…routes to production.ls_test_…routes to staging. Mixing them returns401. - 3
For HTTP integrations — try the API reference
Open the interactive API reference. Paste your key, select environment, fire a live request — no local setup needed.
- 4
For AI agents — connect via MCP
Use the MCP setup wizard or read the MCP guide. Manage keys in MCP settings.