Browse docs▾

MCP Server Integration

Connect auto-social.io tools directly to Cursor, Claude Desktop, and AI agents via Model Context Protocol.

The Model Context Protocol (MCP) allows AI assistants like Cursor, Claude Desktop, and autonomous agents to discover and invoke auto-social.io actions directly.

Remote MCP Endpoint

auto-social.io exposes a standards-compliant Streamable HTTP MCP server:

text
https://auto-social.io/api/mcp

Supported Protocols

  • Transport: Streamable HTTP / Server-Sent Events (SSE) and JSON-RPC 2.0 POST.
  • Authentication: Bearer token via Authorization: Bearer ask_... or OAuth 2.0 PKCE.

Configuring Cursor

To use auto-social.io tools inside Cursor:

  1. Open Cursor Settings, Features, MCP Servers.
  2. Click Add New MCP Server.
  3. Choose HTTP (or SSE) as transport type.
  4. Set URL to https://auto-social.io/api/mcp.
  5. Set authorization header: Authorization: Bearer ask_your_api_key.

Alternatively, configure your project's .cursor/mcp.json:

json
{  "mcpServers": {    "auto-social": {      "url": "https://auto-social.io/api/mcp",      "headers": {        "Authorization": "Bearer ask_your_api_key"      }    }  }}

Configuring Claude Desktop

Add auto-social.io to your claude_desktop_config.json:

json
{  "mcpServers": {    "auto-social": {      "command": "npx",      "args": ["-y", "@auto-social-io/cli", "mcp"]    }  }}

Or configure direct HTTP transport if supported by your version of Claude Code:

bash
claude mcp add --transport http auto-social https://auto-social.io/api/mcp

Available MCP Tools

When connected, your agent receives full schemas and descriptions for all core actions:

  • list_campaigns: Query autopilot campaigns and scheduled cadences.
  • get_campaign_status: Inspect campaign run logs, metrics, and health.
  • pause_campaign / resume_campaign: Manage autopilot execution states.
  • create_campaign / update_campaign: Configure targeting, prompt styles, and frequencies.
  • list_integrations: Retrieve connected social network accounts (Instagram, X, LinkedIn, etc.).
  • preview_publication: Test post renderings and validate platform limits prior to publishing.
  • publish_post: Queue posts for immediate distribution across one or more networks.
  • update_published_post / delete_published_post: Manage live publications.