AI Integration using the MCP Server

Use our MCP server to allow your AI agents to post directly to your social accounts.

With the rise of AI-assisted development, integrating social media capabilities into your workflow has never been faster. By utilizing our Model Context Protocol (MCP) server, your favorite AI assistants can interact directly with Post for Me locally to build and test your integrations.

Here is a breakdown of how our MCP server works and how to set it up for your local coding assistant.

Understanding the MCP Server

Definition: The MCP server provides a standardized way for AI assistants to interact with the Post for Me API and SDK directly within your local development environment.

How it works

Currently, you must run the MCP server locally for it to connect with your AI assistant. Once running, it bridges the gap between the AI's logic and our social media publishing infrastructure.

Installing the Server

Definition: Setting up the server allows seamless integration with popular modern tools and IDEs.

How it works

You can run the MCP Server directly via your terminal using npx -y post-for-me-mcp@latest. Alternatively, there are specific installation configurations available to easily add the server to tools like Claude Code, Cursor, or VS Code.

Requirements: To successfully connect, you must authenticate the server by securely providing your API key as an environment variable (such as POST_FOR_ME_API_KEY or x-post-for-me-api-key).

Operating in Code Mode

Definition: The server utilizes a specific "Code Mode" tool scheme, allowing your AI agent to write actual code against the Post for Me TypeScript SDK.

How it works

The server automatically exposes two specific tools to your agent: a docs search tool for querying general API and SDK documentation, and a code execution tool.

The User Experience: The AI agent writes code and executes it in an isolated sandbox environment that is restricted from web or filesystem access. Because it is sandboxed, it is highly secure. Anything the code returns or prints is sent right back to the agent as a result, allowing your AI to perform complex, multi-step integrations deterministically and repeatably.