bangermapall free tools

A YouTube MCP server on your own API key

bangermap-mcp gives Claude, Cursor or any MCP client the same outlier measurement as the Bangermap app, free, on your own YouTube Data API key. No account, no credits, no server in the middle.

Setup

Create a free YouTube Data API key in the Google Cloud console, two screens and no billing details. Then point your client at the server.

Claude Code, one command.

claude mcp add bangermap --env YOUTUBE_API_KEY=your-key -- npx -y bangermap-mcp

Claude Desktop, in claude_desktop_config.json.

{
  "mcpServers": {
    "bangermap": {
      "command": "npx",
      "args": ["-y", "bangermap-mcp"],
      "env": { "YOUTUBE_API_KEY": "your-key" }
    }
  }
}

Every other MCP client works the same way. The server speaks stdio and needs only YOUTUBE_API_KEY in its environment.

The skill, for Claude Code

The server does the arithmetic. The skill teaches the method, what a baseline is, why the median of settled uploads of the same format, how to read a multiple, and when an outlier is noise. Two commands install it.

/plugin marketplace add AdesiaHQ/bangermap-integrations
/plugin install youtube-outlier-research@bangermap

It follows the open Agent Skills format, so the same folder works in any client that reads skills. The source sits in the same public repo as the server.

Four tools

ToolWhat it answersQuota
find_outliersWhich of this channel's recent uploads beat its own baseline, and by how much~5 units
channel_baselineWhat normal looks like for this channel, so a view count can be judged~5 units
compare_channelsWhat is working across a set of competitors, each measured against itself~5 units per channel
sweep_nicheWhich channels a seed channel is connected to, and what is overperforming on them~5 units per channel

Channels can be given as a URL, an @handle, a channel id, or the URL of any video on the channel. Your free daily quota is 10,000 units, which covers roughly 1,100 scans, and whatever the assistant does not spend stays yours.

The measurement

An outlier is a video that beat its own channel's normal performance. The baseline is the median views of the channel's recent uploads of the same format, counting only uploads old enough to have settled, with Shorts and long-form kept separate. A 10x on a small channel is a stronger signal than a million views on a channel that always gets a million, because it is proof an idea worked without an audience behind it. The same engine runs the free browser tool and the app, so the numbers never disagree.

It is listed in the official MCP registry as io.github.AdesiaHQ/bangermap-mcp, published with a provenance attestation from a public repo, and the n8n community node n8n-nodes-bangermap runs the same operations inside workflows.

Questions

What can Claude do with it?

Ask it which of a channel's uploads beat that channel's own baseline, what normal looks like for a channel, what is working across a set of competitors, or which channels a seed channel is connected to and what is overperforming on them. The server measures, the assistant reasons over the results.

Why does it need my own API key?

Because then nothing is metered and nobody is in the middle. YouTube gives every Google account a free Data API key with 10,000 units a day, a channel scan costs about 5, and requests go from your machine to Google. There is no Bangermap account and no credit allowance.

Which clients does it work with?

Any MCP client that runs stdio servers, including Claude Code, Claude Desktop, Cursor and the rest. It is a normal npm package started with npx.

Is it in the official MCP registry?

Yes, as io.github.AdesiaHQ/bangermap-mcp, published with provenance from a public repo. Directories that syndicate the registry list it automatically.

Does an n8n version exist?

Yes, n8n-nodes-bangermap runs the same four operations inside n8n workflows on the same free key, with filters that turn a scheduled run into an alert, and it is exposed as an AI Agent tool.