M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
F

Finance Toolkit

Community
Transparent, efficient financial analysis
How to install ↓ Compare GitHub source repository ↗
Category
Other #71 of 230
Stars
★ 5.4k Very popular
Transport
Streamable HTTP · stdio (local process)
Runtime
Python · Hosted (no local runtime)
Credentials
API key / credential required
License
MIT
Last commit
60FMRS · C

The supplied sources identify this as a FinanceToolkit server, manifest version 2.1.4, with hosted streamable HTTP and local stdio transports. It emphasizes transparent calculations from raw statements but depends on an FMP API key and external financial-data providers.

Strongest · Documentation 14/20 Weakest · Reliability 10/20

Reliability
10/20
Security and permissions
10/20
Maintenance
13/20
Documentation
14/20
Setup experience
13/20
Why each score
Reliability 10/20
The manifest declares both a streamable-HTTP remote endpoint and a PyPI stdio package, while the README describes the main metrics and source-fallback behavior, making the happy path plausible. Points are deducted because the supplied material contains no CI, committed tests, or evidence for MCP initialization/tool-list handshakes, and it does not establish that manifest version 2.1.4 and package version 2.1.2 behave consistently. Error handling, edge cases, and dependency control are not verifiable.
Security and permissions 10/20
The API key is declared as required and secret, and the hosted connection uses OAuth. The documented functionality is primarily financial-data retrieval and calculation, with no evidence of deletion, payment, deployment, or remote-command actions. Points are deducted because data is sent to Financial Modeling Prep and may fall back to Yahoo Finance, while caching writes pickle files; data boundaries, hosted-service permissions, privacy handling, cache protection, and confirmation controls are insufficiently documented. No red-line credential theft or malicious behavior is evidenced.
Maintenance 13/20
The project states an MIT license, is not archived, provides version identifiers, a GitHub repository, contribution guidance, and maintainer contact details, establishing basic ownership and an update path. Points are deducted because the supplied material provides no commit history, release cadence, dependency-update evidence, vulnerability-response process, or security-advisory channel; stars and open-issue counts are not maintenance proof.
Documentation 14/20
The README covers hosted and local installation, several mainstream clients, OAuth/API-key handling, major modules, examples, free-tier limits, source fallback, caching, benchmarking, and troubleshooting, so documentation is comparatively strong. Points are deducted because there is no reviewable MCP tool inventory, per-tool parameter/return documentation, protocol error behavior, hosted-service pricing/quotas, or complete platform-compatibility detail. The manifest and README/PyPI versions also differ.
Setup experience 13/20
The hosted endpoint offers a direct connection path with instructions for multiple clients. The local path includes a uvx command, manual JSON configuration, and an MCPB bundle, making setup reasonably actionable. Points are deducted because an FMP API key and external-service flow are required; the setup wizard writes client configuration, platform compatibility and recovery from failures are not fully documented, and there is no file-level test or execution evidence.

Static review · not runListed 2026-07-29

Read the FMRS scoring method →

Fit and risk

What it can accessUses the network

Best for

  • Financial researchers and quantitative analysts who need transparent calculation methods
  • Python users performing fundamental, performance, risk, or technical analysis
  • Users who want to query multiple categories of financial data through an MCP client

Not for

  • Users unwilling to provide a Financial Modeling Prep API key
  • Users seeking to bypass data-provider plan or request limits
  • Users requiring a single data source with no Yahoo Finance fallback

Required permissions

  • Requires a Financial Modeling Prep API key; local installations provide it through FINANCIAL_MODELING_PREP_API_KEY
  • Remote connections require completing OAuth and providing an FMP API key on first use
  • The server needs access to external financial data sources for market data, statements, and economic indicators

Risks and side effects

  • Financial data comes from external providers, and the project does not take responsibility for its accuracy
  • Free plans impose limits on daily requests, data history, and exchange coverage
  • API rate limits, network conditions, or invalid keys can make retrieval fail or become slow
  • When Financial Modeling Prep retrieval fails, the default fallback to Yahoo Finance can change the data source

Setup

Before you start

Runtime：Python · Hosted (no local runtime)

FINANCIAL_MODELING_PREP_API_KEY requiredsecret A Financial Modeling Prep API key used to access 30+ years of financial statements; obtain one at https://www.jeroenbouma.com/fmp (free tier available).

Remote: add https://financetoolkit.jeroenbouma.com/mcp to an MCP-compatible client; on first connection, use OAuth to provide a Financial Modeling Prep API key.\nLocal: run uvx --from "financetoolkit[mcp]" financetoolkit-mcp-setup, or use the client configuration in install_config and set FINANCIAL_MODELING_PREP_API_KEY.

claude_desktop_config.json 
{"mcpServers":{"finance-toolkit":{"command":"uvx","args":["--from","financetoolkit[mcp]","financetoolkit-mcp"],"env":{"FINANCIAL_MODELING_PREP_API_KEY":"YOUR_API_KEY_HERE"}}}}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json
{
  "servers": {
    "finance-toolkit": {
      "command": "uvx",
      "args": [
        "--from",
        "financetoolkit[mcp]",
        "financetoolkit-mcp"
      ],
      "env": {
        "FINANCIAL_MODELING_PREP_API_KEY": "YOUR_API_KEY_HERE"
      }
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal
claude mcp add finance-toolkit -e FINANCIAL_MODELING_PREP_API_KEY=YOUR_API_KEY_HERE -- uvx --from 'financetoolkit[mcp]' financetoolkit-mcp

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

After connecting to finance-toolkit in your client's MCP server list, ask something like "What is Apple's latest price-to-earnings ratio?" — a concrete value calculated from financial statements confirms the connection works.

Troubleshooting

  1. Verify that FINANCIAL_MODELING_PREP_API_KEY is set and valid
  2. Check Financial Modeling Prep plan limits, daily request quotas, and rate limits
  3. For remote connections, confirm that the OAuth flow completed and an FMP API key was supplied
  4. Check network connectivity; initializing the Toolkit with multiple tickers can avoid single-ticker thread waiting
  5. To disable fallback, set enforce_source="FinancialModelingPrep"; alternatively select YahooFinance explicitly

Things to try

Once connected, you can ask your AI assistant things like:

  • What are the ROE figures for Apple and Microsoft over the last five years?
  • Show me the extended DuPont analysis for AAPL.
  • List Microsoft's liquidity ratios like current and quick ratio.
  • How have the US unemployment rate and CPI inflation trended in recent years?

Use cases

Query and calculate financial ratios, valuation metrics, and profitability measures
Analyze historical prices, returns, volatility, Sharpe Ratio, and Value at Risk
Retrieve financial statements and run DuPont, WACC, EVA, Altman Z-Score, and related models
Analyze option pricing and Greeks, technical indicators, fixed-income measures, and economic indicators
Use natural language in MCP clients to analyze equities, ETFs, funds, indices, currencies, and cryptocurrencies

Supported clients

Claude Desktop
Claude.ai
Claude Code
VS Code
Cursor
Windsurf

Listed from the project's documentation, not tested by this site.

Overview

Finance Toolkit is an open-source financial-analysis MCP server exposing 200+ metrics, models, and economic indicators calculated from raw financial statements. It supports historical market data, financial statements, ratios, option Greeks, performance and risk metrics, technical indicators, fixed income, economics, portfolio analysis, and instruments including equities, ETFs, funds, indices, currencies, and cryptocurrencies. Financial Modeling Prep is the default data source, with optional fallback to Yahoo Finance when retrieval fails.

Similar servers

BioMCP 75 · B

One binary. One grammar. Evidence from the biomedical sources you already trust.

★ 659 · Tools 11 Compare with this →
Compare with other MCP servers

Source revision e5a3476b72ff Data synced 2026-10-19 Read the FMRS scoring method