MCP (AI Agents)

Postbox exposes an MCP (Model Context Protocol) server that lets AI agents interact with your forms and submissions directly. Requires Pro plan.

What agents can do

Via MCP, an AI agent can:

• List your forms and their schemas
• Read and analyze submissions
• Translate submissions on demand
• Run AI spam analysis
• Generate smart replies
• Get dashboard statistics and submission summaries

Endpoint

https://usepostbox.com/mcp

Transport: Streamable HTTP. Authentication: OAuth 2.1 (Authorization Code + PKCE).

Connecting in Claude.ai / ChatGPT

1. Open MCP client settings
2. Add a new remote MCP server with URL: https://usepostbox.com/mcp
3. The client redirects you to Postbox to authorize
4. Sign in and approve - connected

Connecting in Claude Code, Cursor, or CLI tools

Add to your MCP configuration:

{
  "mcpServers": {
    "postbox": {
      "type": "http",
      "url": "https://usepostbox.com/mcp"
    }
  }
}

The client handles the OAuth flow automatically on first connection.

Claude Desktop

Add to ~/.config/claude-desktop/mcp.json:

{
  "mcpServers": {
    "postbox": {
      "url": "https://usepostbox.com/mcp"
    }
  }
}

Claude will handle the OAuth flow automatically when you first use a Postbox tool.

The auth page looks like this:

OpenCode

Add to ~/.config/opencode/mcp.json:

{
  "mcpServers": {
    "postbox": {
      "url": "https://usepostbox.com/mcp"
    }
  }
}

OpenCode will handle the OAuth flow automatically when you first use a Postbox tool. And if it doesn’t do that, you can run this command:

opencode mcp auth postbox

and then restart OpenCode. Your OpenCode will now have access to your forms, and will look like this (notice the MCP status):

Available tools

Once connected, the agent has access to these tools (all scoped to the authenticated user’s account):

Tool	Description
list_forms	List all your forms
get_form	Get form details including field schema and AI settings
list_submissions	List submissions with optional filtering (inbox, spam, all)
get_submission	Get full details of a single submission
get_dashboard_stats	Get overview statistics (total forms, submissions, spam blocked, AI replies)
translate_submission	Translate a submission on demand (uses AI credits)
analyze_spam	Run AI spam analysis on a submission (uses AI credits)
draft_reply	Generate a reply using a knowledge base (uses AI credits)
summarize_submissions	Get a summary of recent submissions with patterns and insights

Why MCP?

MCP makes Postbox agent-native. Instead of an agent writing code to call your API, it uses structured tools with typed parameters. The agent understands your form schemas and can submit valid data, analyze trends, draft replies, and manage your data - all conversationally.