FeedVector Logo

MCP (Model Context Protocol)

Connect AI agents to FeedVector using the Model Context Protocol (MCP)

Documentation Index

Fetch the complete documentation index at:

https://feedvector.com/llms.txt

Use this file to discover all available pages before exploring further.

Introduction

MCP lets AI agents interact with FeedVector directly through a standardized tool-calling interface. This means you can connect Claude, Cursor, or any MCP-compatible client to your FeedVector account and manage your social scheduling through natural language.

How It Works

FeedVector exposes an MCP server with a set of tools. Your AI client discovers those tools, reads their schemas, and calls them on your behalf.

Available Tools

Tool availability can vary by account and enabled features. Commonly used tools include:

  • integrationList — list connected social media accounts
  • integrationSchema — fetch platform-specific rules/settings schema
  • schedulePostTool — schedule, draft, or publish posts

Authentication

There are two common ways to authenticate with an MCP server:

How to get your API key

  1. Create an account at app.feedvector.com
  2. Connect at least one social media account
  3. Go to Settings > MCP and copy your API key

API Key

Use your API key directly in the MCP endpoint URL or as a Bearer token.

OAuth Token (if applicable)

If you’re building an app for other users and OAuth2 is enabled, use OAuth2 to obtain tokens and pass them the same way as API keys.

Connecting

Use the /mcp endpoint with a Bearer token, or embed your API key in the URL using /mcp/<apiKey>.

Quick Example

A typical interaction looks like:

  1. Agent calls integrationList to see your connected accounts
  2. Agent calls integrationSchema for a platform to learn posting rules
  3. Agent calls schedulePostTool to schedule the post with valid settings
“Schedule a post to LinkedIn for tomorrow at 10am: Excited to announce our new feature!”

Client Setup

Configure your AI client to connect to the FeedVector MCP server.

Claude Desktop

Add the following to your Claude Desktop MCP configuration file:

macOS

Edit ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "feedvector": {
      "url": "https://app.feedvector.com/api/mcp/your-api-key"
    }
  }
}

Windows

Edit %APPDATA%\\Claude\\claude_desktop_config.json:

{
  "mcpServers": {
    "feedvector": {
      "url": "https://app.feedvector.com/api/mcp/your-api-key"
    }
  }
}

Replace your-api-key with your actual API key from Settings > Public API.

Claude Code

Add FeedVector as an MCP server in your Claude Code configuration:

{
  "mcpServers": {
    "feedvector": {
      "url": "https://app.feedvector.com/api/mcp/your-api-key"
    }
  }
}

Cursor

In Cursor, go to Settings > MCP and add a new server:

  • Name: FeedVector
  • Type: HTTP
  • URL:
    https://app.feedvector.com/api/mcp/your-api-key

Other MCP Clients

Any MCP-compatible client can connect to FeedVector. Use the streamable HTTP transport:

  • URL:
    https://app.feedvector.com/api/mcp/your-api-key
  • Transport: Streamable HTTP

Or, if your client supports Bearer token authentication:

  • URL:
    https://app.feedvector.com/api/mcp
  • Transport: Streamable HTTP
  • Authorization:
    Bearer your-api-key

Verify Connection

Once connected, ask your AI agent:

“List my connected social media accounts”

If the connection is working, the agent will call the integrationList tool and return your connected accounts.