# polygon-io/mcp_polygon) [Health: Active]

**Category:** 💰 Finance & Fintech  
**Repository:** https://github.com/polygon-io/mcp_polygon  
**GitHub Stars:** 390  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/polygon-io-mcp-polygon-

## Description
) 🐍 ☁️ -  An MCP server that provides access to Polygon.io financial market data APIs for stocks, indices, forex, options, and more.

## Tools
Capabilities this server exposes over MCP:

- **search_endpoints** — Search for API endpoints and built-in functions by natural language query. Returns titles, path patterns, and descriptions. Set `detail` to `"more"` for query parameter docs, or `"verbose"` for full documentation. Use `max_results` to limit results.
- **call_api** — Call any Massive.com REST API endpoint. Supports storing results as an in-memory database table (`store_as`) and applying post-processing functions (`apply`). Paginated responses include a next-page hint.
- **query_data** — Run SQL against stored SQLite DB. Supports `SHOW TABLES`, `DESCRIBE <table>`, `DROP TABLE <table>`, CTEs, window functions, and more. Results can also be post-processed with `apply`.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `uvx` (confidence: high):

```json
"mcpServers": {
  "mcp-polygon": {
    "command": "uvx",
    "args": ["--from"],
    "env": {
      "MASSIVE_API_KEY": ""
    }
  }
}
```

**Requires environment variables:** `MASSIVE_API_KEY` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What polygon-io/mcp_polygon) MCP server does

The polygon-io/mcp_polygon) MCP server provides an MCP interface to the full Massive.com financial data API. Its data coverage includes stock, options, forex, crypto, and futures aggregates; historical and real-time trades and quotes; snapshots; ticker reference data; dividends; splits; IPOs; fundamentals; analyst ratings; news; treasury yields; inflation data; and market status information.

Rather than defining a separate MCP tool for every financial endpoint, it exposes a smaller set of composable operations. An agent can discover an endpoint, call it with the required parameters, save the response for later analysis, and query the saved data with SQL.

## How it works

`search_endpoints` searches the dynamically indexed API documentation and built-in functions using natural language. Results can include endpoint titles, path patterns, descriptions, query parameter details, or full documentation depending on the requested detail level. The index is loaded from Massive.com's `llms.txt` document at startup.

`call_api` sends requests to Massive.com REST endpoints. Responses may be stored as in-memory SQLite tables with `store_as`, and post-processing can be applied through the `apply` parameter. Paginated responses provide a hint for retrieving the next page.

`query_data` executes SQL against stored tables. Supported operations include table inspection and deletion, common table expressions, window functions, and other SQLite queries. Results can also be passed through built-in post-processing functions.

## Setup and configuration

The polygon-io/mcp_polygon) MCP server requires Python 3.12 or newer, Astral UV 0.4.0 or newer, and a Massive.com API key. The documented installation uses UV to install the `mcp_massive` package from the project's Git repository. Claude Desktop can then launch the installed `mcp_massive` binary over the default STDIO transport.

Set `MASSIVE_API_KEY` to authenticate requests. `POLYGON_API_KEY` remains available as a deprecated alias. Optional settings can change the transport, API base URL, endpoint-index URL, maximum number of in-memory tables, and maximum rows per stored table. Supported transports are STDIO, SSE, and streamable HTTP; STDIO is the default.

## Tools and capabilities

- Search API endpoints and built-in functions with natural-language queries.
- Call any supported Massive.com REST endpoint.
- Store API responses in temporary SQLite tables.
- Inspect, query, and remove stored tables with SQL.
- Apply Black-Scholes price and Greek calculations, including delta, gamma, theta, vega, and rho.
- Calculate simple, logarithmic, and cumulative returns, Sharpe ratio, Sortino ratio, SMA, and EMA.

## Limitations and notes

The polygon-io/mcp_polygon) MCP server is marked experimental, so breaking changes may occur. It depends on access to the Massive.com API and its applicable terms and privacy policy. Stored data is held in an in-memory SQLite database, with configurable limits of 50 tables and 50,000 rows per table by default. The README states that the server does not store or cache user data and proxies requests to Massive.com.

_Full upstream README: https://allmcps.com/mcp/polygon-io-mcp-polygon-/readme_