← Back to blog

Connect ReelsFarm to Your AI Agent Workflow — The Full Setup Guide

· MCP & AI Agents · 6 min read

ReelsFarm connects to AI agents through two paths. An MCP server for Claude Code, Cursor, Codex, and any MCP-compatible client. A CLI and TypeScript SDK for terminal-native workflows and custom integrations. Here is how to set up both and start generating content from your agent.

ReelsFarm gives your AI agent two integration paths.

MCP server for protocol-based tool discovery. CLI and SDK for programmatic control.

Here is how to set up both.

Choose Your Path

Use the MCP server if you use Claude Code, Cursor, Codex, or any MCP-compatible AI coding agent. The agent discovers tools automatically. You get 45+ tools for generation, scheduling, publishing, and account management.

Use the CLI and SDK if you are building a custom agent, automating from CI/CD, or prefer terminal commands. The CLI is `reelsfarm`. The SDK is `@reelsfarm/mcp-client`. Both are open source under MIT.

You can use both. The MCP server for agent conversations. The CLI for scripts and automation.

MCP Server Setup

Create an account at reelsfarm.com with a paid plan. MCP server access and Agent SDK are included on all plans.

Go to account settings. Open the MCP tab. Create an API key. Copy it immediately.

Add the ReelsFarm MCP server to your MCP client config file with your API key. The server runs over HTTP. No local process to manage.

Restart your AI coding agent. It discovers ReelsFarm tools through the MCP protocol on startup.

Verify by asking your agent to list your connected social accounts.

CLI Setup

Install the CLI globally or as a project dependency.

```bash npm install -g @reelsfarm/mcp-client ```

Authenticate with your API key.

```bash reelsfarm login --api-key rfmcp_xxx ```

Verify the connection.

```bash reelsfarm whoami ```

The CLI and SDK are open source. The generation infrastructure behind them is paid.

Your First Content

From an MCP-connected agent, ask it to create content.

Describe what you want. The agent uses ReelsFarm tools to generate drafts. You review each draft. You approve. Content publishes or schedules. Every destructive action waits for confirmation.

From the CLI, use commands directly.

```bash reelsfarm avatars generate --prompt "creator selfie style" reelsfarm slideshows prepare --product "Summer Serum" --format tiktok --slides 6 reelsfarm posts list --json ```

Same capability. Different interface.

What Happens Next

Your agent becomes a content creator. It generates UGC, creates slideshows, manages AI characters, and publishes to TikTok, Instagram, YouTube, and Facebook.

You stay in the review layer. The agent executes. You approve. Content goes live.

No dashboard. No compose box. No context switching. Social media content creation from the same environment you write code in.

Related reading

Related comparisons

Turn one idea into a week of content.

Create, schedule, and publish AI-powered posts from one workflow built for consistent social growth.

Start for free