Channel3 Ai Mcp Server

Channel3 MCP Server

Quick Start (Free Tier)

No API key required. Connect directly and start searching:

https://mcp.trychannel3.com/

The free tier is rate-limited to 10 requests per minute per IP address. If you hit the limit, you'll receive an error response.

Unlimited Access

For unlimited usage and affiliate tracking, create an account and add your API key:

1. Create an account at trychannel3.com and create an API key to authenticate requests and affiliate purchases to your account.
2. Add the Channel3 MCP to your agent with ?apiKey= in the URL as such:

https://mcp.trychannel3.com/?apiKey=placeholder

Cursor

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "Channel3": {
      "url": "https://mcp.trychannel3.com/?apiKey="
    }
  }
}

To use the free tier (no API key, rate-limited), omit the apiKey parameter:

{
  "mcpServers": {
    "Channel3": {
      "url": "https://mcp.trychannel3.com/"
    }
  }
}

VS Code

Add to .vscode/mcp.json:

{
  "servers": {
    "Channel3": {
      "type": "http",
      "url": "https://mcp.trychannel3.com/?apiKey="
    }
  }
}

Claude Code

claude mcp add --transport http Channel3 "https://mcp.trychannel3.com/?apiKey="

Claude Desktop

Add to your Claude Desktop config file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "Channel3": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.trychannel3.com/sse?apiKey="
      ]
    }
  }
}

OpenAI Agents SDK (Python)

import asyncio
from agents import Agent, Runner
from agents.mcp import MCPServerStreamableHttp

async def main():
    async with MCPServerStreamableHttp(
        name="Channel3",
        params={
            "url": "https://mcp.trychannel3.com/?apiKey=",
        },
        cache_tools_list=True,
    ) as server:
        agent = Agent(
            name="Shopping Agent",
            instructions="You are a personal shopping assistant.",
            mcp_servers=[server],
        )
        result = await Runner.run(agent, "I'm looking for a new laptop")
        print(result.final_output)

asyncio.run(main())

OpenAI Agents SDK (TypeScript)

import { Agent, run, MCPServerStreamableHttp } from "@openai/agents";

const server = new MCPServerStreamableHttp({
  url: "https://mcp.trychannel3.com/?apiKey=",
  name: "Channel3",
});

await server.connect();
const agent = new Agent({
  name: "Shopping Agent",
  instructions: "You are a personal shopping assistant.",
  mcpServers: [server],
});
const result = await run(agent, "I'm looking for a new laptop");
console.log(result.finalOutput);
await server.close();

Local Testing

1. Start the dev server: pnpm run dev
2. Start MCP Inspector: npx @modelcontextprotocol/inspector
3. In the MCP Inspector UI:

• Transport Type: Streamable HTTP
• URL: http://localhost:8787/?apiKey=
• Click Connect

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: channel3-ai
• Source: channel3-ai/mcp-server
• License: MIT

Install and usage instructions live in the source repository linked above.