mcpField
C
cryptoetf

cryptoetf

Remote
cryptoetf.todayFinance0Updated 2w ago

About cryptoetf

cryptoetf is an MCP data service for spot crypto ETF flows, fund details, the CEFI index, and live prices across 16 assets. It derives flow figures from issuers’ daily reports and supports remote MCP clients.

Key features

  1. 01Latest ETF flows for 16 assets
  2. 0230-day asset and fund flow history
  3. 03Fund directory and AUM data
  4. 04Seven-day flow analytics
  5. 05CEFI index and live coin prices

Use cases

  1. 01Check daily ETF net flows by asset
  2. 02Compare monthly BTC and ETH flows
  3. 03Find the asset with the strongest weekly inflows

MCP support

It provides an unauthenticated remote MCP server with tools for flows, fund data, the CEFI index, and prices. A stdio bridge is also available for clients that only support stdio.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://mcp.cryptoetf.today/api/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Packages
npm@zerosix-studio/cryptoetf-mcp1.0.12

Tools

7

This tool takes no parameters.

Details

From the project README

CryptoETF MCP — spot crypto ETF flows for AI agents

npm npm downloads MCP Registry

The Model Context Protocol surface of cryptoetf.today: daily net flows for every US-listed spot crypto ETF complex and every fund in it, fund AUM, the CEFI sentiment index and live coin prices — for Claude, Cursor, Codex, Gemini CLI and any other MCP-capable agent.

Sixteen assets: BTC, ETH, SOL, XRP, HYPE, DOGE, LINK, AVAX, HBAR, LTC, BNB, DOT, SUI, NEAR, TRX, ZEC. Flow figures are net USD millions, sourced from the issuers' own daily reports; for the eleven smaller complexes there is no market-wide aggregator, so the numbers are computed from each issuer's fund pages.

You probably don't need to install anything. The server is hosted, and every client that speaks remote MCP can reach it directly:

https://mcp.cryptoetf.today/api/mcp     Streamable HTTP · JSON-RPC 2.0

No API key, no account — the endpoint is open.

Setup per client

Claude Code — one command, no restart:

claude mcp add --transport http cryptoetf https://mcp.cryptoetf.today/api/mcp

Cursor — .cursor/mcp.json (project) or ~/.cursor/mcp.json (global):

{
  "mcpServers": {
    "cryptoetf": { "url": "https://mcp.cryptoetf.today/api/mcp" }
  }
}

Claude Desktop and other stdio-only clients — this package is the bridge. claude_desktop_config.json, then restart the app fully:

{
  "mcpServers": {
    "cryptoetf": {
      "command": "npx",
      "args": ["-y", "@zerosix-studio/cryptoetf-mcp"]
    }
  }
}

Nothing to install by hand: npx fetches the package on first run. It has zero dependencies and simply forwards JSON-RPC between stdin/stdout and the hosted endpoint.

Tools

ToolWhat it returns
get_flows_summaryLatest-day net flow for all 16 assets, with the date each figure belongs to
get_asset_flowsDaily net flow history for one asset over the last 30 days
get_fund_flowsThe same 30 days for one asset, fund by fund (IBIT, FBTC, ETHA…), with ticker and issuer
get_fundsFund directory: issuer, fee, listing date, current net assets (AUM) and coins held
get_weekly_analytics7-day totals per asset: net flow, average day, positive/negative day counts
get_cefi_indexComposite CEFI index (0–100, 50 = zero net flow) with its flow, trend and breadth components
get_pricesSpot price and 24h change for every tracked asset

History reaches back 30 days, in the tools and in the REST API alike. Longer history is on cryptoetf.today.

Ask it things like

  • "What are today's crypto ETF flows?"
  • "Compare BTC and ETH ETF flows over the past month"
  • "Which asset had the strongest week of inflows?"
  • "Is the CEFI index bullish right now?"

Configuration

VariableDefaultPurpose
CRYPTOETF_MCP_URLhttps://mcp.cryptoetf.today/api/mcpPoint the bridge at another endpoint
CRYPTOETF_API_KEY—Sent as Authorization: Bearer … when set; the public endpoint does not need it

Requires Node.js 18 or newer.

Data and limits

Flow data is updated every business day as issuers publish. Figures are informational — this is market data, not investment advice.

Privacy Policy

Full policy: cryptoetf.today/en/privacy.

What is collected. The bridge forwards your JSON-RPC requests to https://mcp.cryptoetf.today/api/mcp and returns the responses. The only request data that leaves your machine is the tool name and its arguments (an asset ticker, at most). Conversation content, files and credentials are never read or transmitted.

How it is used and stored. The server logs the standard request metadata of any web service — IP address, timestamp, endpoint — and keeps a per-IP counter in Redis to enforce the 60 requests/minute rate limit. Logs are retained for 30 days and are used only to operate and protect the service.

Third-party sharing. None. Request data is not sold, shared or passed to any third party. The service is operated by us and calls only our own API.

Retention. Rate-limit counters expire within a minute; request logs within 30 days. No user accounts and no personal profiles exist — the endpoint is open and unauthenticated.

Contact. github.com/zerosix-studio/cryptoetf-mcp/issues or the contact form at cryptoetf.today.

Releasing

Bump the version in package.json, then tag the commit:

git tag v1.0.4 && git push --tags

The Publish to npm workflow runs the bridge against the live server, skips the publish if that version is already in the registry, and otherwise pushes it with provenance — the npm page then links back to the exact workflow run and commit the tarball was built from.

MIT licensed.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.