sifting/io
Remote MCP server · Streamable HTTP

Model Context Protocol (MCP).

Give your AI assistant live market data through the hosted SiftingIO MCP server. Add one URL and your API key to Claude, Cursor, VS Code, or any client that speaks the Model Context Protocol, then ask for prices, fundamentals, filings, historical bars, and on-chain data across stocks, forex, crypto, and commodities in plain language.

View on GitHub

What it does

The SiftingIO MCP server lets your AI assistant pull real market data while it works. It speaks the Model Context Protocol, an open standard that connects assistants to tools, and exposes SiftingIO as a set of those tools. Connect it once and the assistant can fetch live prices, historical bars, fundamentals, filings, market hours, and on-chain data across stocks, forex, crypto, and commodities, all from your own API key.

Hosted, no setup
The server runs at https://mcp.sifting.io/mcp. There is nothing to deploy. You add one URL and your key to your client and start asking questions.

What you'll need

Connecting takes two things:

  • An AI client that supports MCP, such as Claude, Cursor, or VS Code.
  • A SiftingIO API key. Create one from your dashboard.
Step 1

Get your SiftingIO API key

From your dashboard, create an API key. It looks like sft_... and is what the assistant uses to authenticate. Keep it private, the same way you would any password.

Step 2

Add the server to your AI client

In a client that supports remote MCP servers, add a new server with the URL https://mcp.sifting.io/mcp and an X-API-Key header set to your key. Many clients accept a config block like this:

mcp client config
{  "mcpServers": {    "siftingio": {      "url": "https://mcp.sifting.io/mcp",      "headers": { "X-API-Key": "sft_..." }    }  }}
Where the key goes
The key travels as a request header, so it stays out of your prompts and chat history. An Authorization: Bearer sft_... header works too if your client prefers that field.
Step 3

No remote support? Use the bridge

Some clients can only launch a local command, not add a remote URL. For those, mcp-remote runs a small local bridge that forwards to the hosted server and passes your key:

bridge
npx mcp-remote https://mcp.sifting.io/mcp \
  --header "X-API-Key: sft_..."
When to use it
Reach for the bridge only if your client lacks native remote-server support. If the config in the previous step worked, you do not need this.

What the assistant can reach

Once connected, the assistant picks the right tool for each question. The server covers:

Capability
Live prices
Latest price and top-of-book quote across stocks, forex, crypto, and commodities
Historical bars
OHLCV history for stocks, forex, crypto, and commodities
Discovery
Search tickers and pull a company profile
Fundamentals
Financial ratios and individual reported concepts
Filings
List a company's recent regulatory filings
Market status
Whether a market is open, plus trading hours
Economic calendar
Upcoming US economic events
On-chain
Wallet portfolio across supported chains

See the repository for the full tool list and the parameters each one takes.

Step 4

Ask for market data in plain language

You do not call tools yourself. Ask a normal question and the assistant chooses and runs the right one. For example:

  • “What is the current price of Bitcoin?”
  • “Show me the latest financial ratios for Apple.”
  • “List the most recent filings for Microsoft.”
  • “What US economic events are coming this week?”
Symbols
Markets use a USD-quoted form, for example BTCUSD for crypto and EURUSDfor forex. The assistant usually maps a name like “Bitcoin” to the right symbol on its own.

Prefer to run it locally? Use the Python server

Alongside the hosted server, an open-source Python server runs on your own machine, which suits Claude Desktop and local agent setups. It connects over stdio by default and reads your key from an environment variable:

claude desktop config
{  "mcpServers": {    "siftingio": {      "command": "siftingio-mcp",      "env": { "SIFTING_API_KEY": "sft_..." }    }  }}
  • Set SIFTING_API_KEY to your key so the server can authenticate.
  • It also offers a Streamable HTTP mode for remote setups, with an optional auth token.
  • The Python repository has install steps and the full tool list.

Notes and limits

  • Every tool call hits the SiftingIO API and counts against your plan. A single question can use several calls.
  • The hosted server is stateless and authenticates per request, so your key is sent with every call and nothing is stored server-side.
  • Prices are an aggregated fair-price reference computed across sources, consistent with the REST, WebSocket, and FIX products.
  • Both servers are open source under the MIT license. One-click OAuth connect is next on the roadmap.

FAQ

  • What is MCP and what does this server do?

    The Model Context Protocol is an open standard that lets AI assistants call external tools. The SiftingIO MCP server exposes market data as those tools, so an assistant like Claude or Cursor can fetch live prices, fundamentals, filings, and historical bars while it answers you.

  • Which AI clients does it work with?

    Any client that supports remote MCP servers, including Claude, Cursor, and VS Code. Clients without native remote support can connect through the mcp-remote bridge, and a local Python server is available for setups that prefer to run it themselves.

  • Do I need to host or deploy anything?

    No. The hosted server runs at mcp.sifting.io, so you only add its URL and your API key to your client. The open-source repository is there if you want to read the code or run your own copy.

  • How is the connection authenticated?

    With your SiftingIO API key, sent as an X-API-Key header on each request (an Authorization Bearer header is also accepted). The server holds no session state, so the key travels with every call and nothing is stored server-side.

  • What if my client cannot add a remote server directly?

    Use the mcp-remote bridge. It runs a small local command that forwards to the hosted server and passes your API key, so any MCP client that launches a local command can still connect.

  • Is there a Python version?

    Yes. The Python server runs locally over stdio (or Streamable HTTP) and is handy for Claude Desktop and local agent setups. It reads your key from the SIFTING_API_KEY environment variable.

  • Does every question use my API quota?

    Yes. Each tool call the assistant makes hits the SiftingIO API and counts against your plan, so a single question that pulls several data points uses several calls.

Get started

Give your assistant market data

Grab an API key, add the server to your client, and ask your first question in minutes. The repositories have the full tool list and setup notes.

Get an API keyView the server on GitHubPython versionAPI reference