@dropreply/mcp

DropReply MCP Server

MCP (Model Context Protocol) server that lets AI agents publish replies on Reddit and Twitter/X through DropReply.

Installation

Run directly with npx:

npx dropreply-mcp

Or install globally:

npm install -g dropreply-mcp

Configuration

Set your API key as an environment variable:

export DROPREPLY_API_KEY=your_api_key_here

Optionally override the base URL (defaults to https://api.dropreply.com):

export DROPREPLY_BASE_URL=https://api.dropreply.com

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "dropreply": {
      "command": "npx",
      "args": ["-y", "dropreply-mcp"],
      "env": {
        "DROPREPLY_API_KEY": "your_api_key_here"
      }
    }
  }
}

Claude Code

claude mcp add dropreply -- npx -y dropreply-mcp

Then set the env var DROPREPLY_API_KEY in your shell before launching Claude Code.

Available Tools

Tool	Description
publish_reply	Publish a reply/comment on Reddit or Twitter/X using an aged managed account
check_reply_status	Check the status of a previously published reply
upvote	Upvote a Reddit post/comment or like a tweet
check_usage	Check current API usage and remaining credits
list_replies	List recent replies with their status and health

publish_reply

Publishes a reply to a Reddit post or tweet.

• platform (required): reddit or twitter
• target_url (required): URL of the post/tweet to reply to
• content (required): Reply text, 10-2000 characters, no URLs allowed
• schedule_at (optional): ISO 8601 datetime to schedule publishing (Growth/Scale plans only, up to 30 days ahead)

check_reply_status

Checks the current status of a reply by its ID.

• reply_id (required): The ID returned by publish_reply

upvote

Upvotes a Reddit post/comment or likes a tweet.

• platform (required): reddit or twitter
• target_url (required): URL of the target

check_usage

Returns current plan info and credit usage. No parameters.

list_replies

Lists recent replies with optional filters.

• platform (optional): Filter by reddit or twitter
• status (optional): Filter by status (e.g. queued, scheduled, published, failed)
• limit (optional): Number of results (default: 10)