Tangerine Index

npx add-mcp 'https://tangerineindex.com/mcp'

Use it from your AI assistant

Ask “is there already a tool that does X?” from Claude, ChatGPT, Cursor, VS Code, or any assistant that supports MCP. It searches the same products as this site.

MCP server URL

https://tangerineindex.com/mcp

No sign-in, free. Calls are rate-limited.

5 tools

• search_products: Find products by what you need. Can filter by form and open source.
• trending_products: List what drew the most reaction in the last 3, 7, or 30 days. Can narrow by need, form, and open source.
• new_products: List products added to the index within 7 days of their first reaction, most recently added first. Can narrow by need, form, open source, and since.
• get_product: Get one product with all of its reactions, plus its MCP server, Homebrew, and package names when known.
• similar_products: Find products with a similar description.

Set up your client

Claude (web and desktop app)

1. Open Customize › Connectors.
2. Click “+ Add”, then “Add custom connector”.
3. Name it “Tangerine Index”, paste the URL above, and click “Continue”.
4. Under Authentication, choose “No sign in”, then click “Add”.
5. In a chat, click “+” › Connectors and turn on Tangerine Index.

Free plans can add one custom connector. On Team and Enterprise, an owner adds it under Organization settings › Connectors.

Official guide

Claude Code

Run this in your terminal:

claude mcp add --transport http tangerine-index https://tangerineindex.com/mcp

Add --scope user to use it in every project. Check the connection with claude mcp list, or /mcp inside Claude Code.

Or share it with your team in.mcp.json:

{
  "mcpServers": {
    "tangerine-index": {
      "type": "http",
      "url": "https://tangerineindex.com/mcp"
    }
  }
}

Keep "type": "http". Without it, the entry is read as a local (stdio) server and won't connect.

Official guide

ChatGPT

1. Open Plugins at chatgpt.com/plugins.
2. Click “+”, then “Add custom MCP server”.
3. Name it “Tangerine Index”. Under Connection, choose the public endpoint and paste the URL above.
4. Choose no authentication. Click “I understand and want to continue”, then “Create as a plugin”.

Whether you can add custom MCP servers depends on your plan and workspace settings. In a work workspace, an admin may need to allow it.

Official guide

Codex (CLI, IDE extension, ChatGPT desktop app)

Run this in your terminal:

codex mcp add tangerine-index --url https://tangerineindex.com/mcp

Or add it to ~/.codex/config.toml:

[mcp_servers.tangerine-index]
url = "https://tangerineindex.com/mcp"

The Codex CLI, the IDE extension, and the ChatGPT desktop app share this file, so adding it once is enough.

In the IDE extension, open the gear menu › MCP servers › Add server. Choose “Streamable HTTP”, paste the URL above, save, then click “Restart extension”.

Official guide

Cursor

Opens Cursor. If nothing happens, add the JSON below by hand.

Add this to ~/.cursor/mcp.json (all projects) or.cursor/mcp.json (this project):

{
  "mcpServers": {
    "tangerine-index": {
      "url": "https://tangerineindex.com/mcp"
    }
  }
}

Turn it on or off from Customize in the sidebar.

Official guide

VS Code

Opens VS Code. If nothing happens, use one of the steps below.

Or open the Command Palette, run “MCP: Add Server”, choose “HTTP”, paste the URL above, name it tangerine-index, and choose “Global”.

Or add it to mcp.json in your user profile:

{
  "servers": {
    "tangerine-index": {
      "type": "http",
      "url": "https://tangerineindex.com/mcp"
    }
  }
}

VS Code uses the "servers" key. AI features must be turned on.

Official guide

GitHub Copilot CLI

Run this in your terminal:

copilot mcp add --transport http tangerine-index https://tangerineindex.com/mcp

Or run /mcp add inside Copilot CLI, choose HTTP, paste the URL above, and press Ctrl+S. Servers are saved in ~/.copilot/mcp-config.json. For Copilot in VS Code, see VS Code above.

Official guide

GitHub Copilot coding agent

In your repository, open Settings › Copilot › MCP servers, paste this, and click “Save MCP configuration”:

{
  "mcpServers": {
    "tangerine-index": {
      "type": "http",
      "url": "https://tangerineindex.com/mcp",
      "tools": [
        "search_products",
        "trending_products",
        "new_products",
        "get_product",
        "similar_products"
      ]
    }
  }
}

Only repository admins can change this. Every tool only reads data, so listing them by name is safe.

Official guide

Devin Desktop (formerly Windsurf)

Add this to ~/.config/devin/mcp_config.json (all projects) or.devin/mcp_config.json (this project):

{
  "mcpServers": {
    "tangerine-index": {
      "url": "https://tangerineindex.com/mcp"
    }
  }
}

Then refresh the list under Customize › MCPs. Older Windsurf versions read ~/.codeium/windsurf/mcp_config.json and use the key "serverUrl".

Official guide

Antigravity (and Gemini CLI)

Add this to ~/.gemini/config/mcp_config.json (all projects) or.agents/mcp_config.json (this workspace):

{
  "mcpServers": {
    "tangerine-index": {
      "serverUrl": "https://tangerineindex.com/mcp"
    }
  }
}

In the IDE, open “…” at the top of the agent panel › MCP Servers › Manage MCP Servers › View raw config. Type /mcp to check the connection.

Still on Gemini CLI? Run:

gemini mcp add --transport http tangerine-index https://tangerineindex.com/mcp

Add -s user to use it in every project, and check it with gemini mcp list.

Official guide

Other clients

Most clients that support remote MCP only need the URL above. Choose Streamable HTTP (sometimes just “HTTP”) and no authentication.

The field names differ by client. A wrong name is ignored without an error:

• "url": most clients, including Cursor, Codex, Copilot, and Devin Desktop
• "serverUrl": Antigravity and older Windsurf
• "servers": VS Code, for the list of servers
• "type": "http": Claude Code and Copilot, next to "url". Without it, the entry is read as a local server
• "type": "streamableHttp": Cline. Without it, Cline uses SSE

If your client only runs local (stdio) servers, bridge it with mcp-remote:

{
  "mcpServers": {
    "tangerine-index": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://tangerineindex.com/mcp"
      ]
    }
  }
}

mcp-remote is a third-party package and needs Node.js.

Official guide

If it doesn't connect

• Check that the URL ends in /mcp and the transport is Streamable HTTP. SSE and stdio won't connect.
• Check the field names for your client under “Other clients”.
• “Rate limit exceeded” (HTTP 429) means too many calls in a minute. Wait a minute and try again.
• If an install button does nothing, the app may be missing or out of date. Add the JSON by hand.
• Opening the URL in a browser brings you back to this page. That's expected: the URL is for AI assistants.

Try asking

• Is there already an open-source tool that turns PDFs into Markdown?
• Find me a local dictation app for macOS, and show what people said about it.
• What's similar to Langfuse?

You can ask in any language. The search itself works best in English.

Also listed on

• Official MCP Registry (as com.tangerineindex/tangerine-index)
• Glama