mcp_massive MCP — Cas d'usage, Installer & Démo en direct

Pourquoi l'utiliser

Fonctionnalités clés

Dynamic endpoint discovery — no hardcoded tool list for 100+ API endpoints
DataFrame storage with in-memory SQL — chain calls without token explosion
Built-in: Black-Scholes, return calcs, technical indicators
Covers stocks, options, forex, crypto, futures, real-time trades, fundamentals
Install via uv tool install; Python 3.12+

Démo en direct

Aperçu en pratique

massive.replay ▶ prêt

0/0

Installer

Choisissez votre client

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "massive": {
      "command": "uvx",
      "args": [
        "mcp_massive"
      ],
      "_inferred": true
    }
  }
}

Ouvrez Claude Desktop → Settings → Developer → Edit Config. Redémarrez après avoir enregistré.

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "massive": {
      "command": "uvx",
      "args": [
        "mcp_massive"
      ],
      "_inferred": true
    }
  }
}

Cursor utilise le même schéma mcpServers que Claude Desktop. La config projet l'emporte sur la globale.

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "massive": {
      "command": "uvx",
      "args": [
        "mcp_massive"
      ],
      "_inferred": true
    }
  }
}

Cliquez sur l'icône MCP Servers dans la barre latérale Cline, puis "Edit Configuration".

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "massive": {
      "command": "uvx",
      "args": [
        "mcp_massive"
      ],
      "_inferred": true
    }
  }
}

Même format que Claude Desktop. Redémarrez Windsurf pour appliquer.

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "massive",
      "command": "uvx",
      "args": [
        "mcp_massive"
      ]
    }
  ]
}

Continue utilise un tableau d'objets serveur plutôt qu'une map.

~/.config/zed/settings.json

{
  "context_servers": {
    "massive": {
      "command": {
        "path": "uvx",
        "args": [
          "mcp_massive"
        ]
      }
    }
  }
}

Ajoutez dans context_servers. Zed recharge à chaud à la sauvegarde.

claude mcp add massive -- uvx mcp_massive

Une seule ligne. Vérifiez avec claude mcp list. Supprimez avec claude mcp remove.

Cas d'usage

Usages concrets : mcp_massive

Analyze an options chain and price theoretical value

👤 Options traders, quant hobbyists ⏱ ~20 min advanced

Quand l'utiliser : You want to look at AAPL options, compare IV to historical vol, and spot-check Black-Scholes values.

Prérequis

Massive.com API key — massive.com signup — free tier

Déroulement

Find the right endpoint

Search endpoints: 'options chain for a given ticker and expiry'.✓ Copié

→ Best-match endpoint + params
Pull chain

Call the endpoint for AAPL with the next monthly expiry. Store as DataFrame 'aapl_chain'.✓ Copié

→ DataFrame stored; head preview
Analyze

Run SQL: SELECT strike, iv, bid, ask FROM aapl_chain WHERE expiry=... ORDER BY strike. Price Black-Scholes for each and show deltas to market mid.✓ Copié

→ Table of strikes with BS-vs-market

Résultat : Hands-on options screen without writing Python yourself.

Pièges

Delayed data on free tier — Check data timestamps; paid tier for real-time

Compute total return for a portfolio over a period

👤 Investors tracking performance ⏱ ~15 min intermediate

Quand l'utiliser : Year-end — how did my 10-position portfolio perform vs SPY?

Déroulement

Fetch prices

For each ticker in [list], get daily close prices from Jan 1 to today. Store each as a DataFrame.✓ Copié

→ Price series per ticker
Compute weighted return

Using my weights [paste], compute weighted portfolio return. Compare to SPY return over the same period.✓ Copié

→ Portfolio YTD vs SPY YTD

Résultat : Quick portfolio report without spreadsheets.

Pièges

No dividend reinvestment accounting by default — Use adjusted-close series; ask Claude to note the assumption

Combiner avec : filesystem

Screen stocks on fundamental criteria

👤 Value investors ⏱ ~20 min intermediate

Quand l'utiliser : You want US large-cap stocks with P/E < 15, dividend yield > 3%, and positive 5y earnings growth.

Déroulement

Find endpoints

Search endpoints for 'fundamentals' and 'screener'. Pick the right one.✓ Copié

→ Endpoint choice
Screen

Pull fundamentals for S&P 500 components. Store as DataFrame. Run SQL to filter my criteria.✓ Copié

→ Matching tickers with key ratios

Résultat : A shortlist to research further.

Pièges

Fundamentals data lags quarterly filings by weeks — Check as_of dates

Combinaisons

Associez-le à d'autres MCPs pour un effet X10

massive + filesystem

Persist analysis to CSV for Excel follow-up

After the screen, export the final DataFrame as /reports/screen-YYYY-MM-DD.csv.✓ Copié

Outils

Ce que ce MCP expose

Outil	Entrées	Quand appeler	Coût
search_endpoints	query: str	First step — discover what's available	free
get_endpoint_docs	endpoint_id	Before calling an endpoint you don't know	free
call_api	endpoint, params, store_as?: str	Actually fetch data; store to re-query	1 API call
query_data	sql: str	Analyze stored DataFrames without re-fetching	free

Coût et limites

Coût d'exécution

Quota d'API: Per Massive.com plan
Tokens par appel: Raw series can be large — use DataFrame + SQL to slice
Monétaire: Free tier exists; paid for real-time / higher volume
Astuce: store_as + query_data is dramatically cheaper than re-fetching and re-parsing each time.

Sécurité

Permissions, secrets, portée

Stockage des identifiants : MASSIVE_API_KEY env var

Sortie de données : Queries to massive.com

Project is marked experimental; breaking API changes possible. Pin version.
Financial data is for info only — this is not investment advice and the LLM can miscompute.

Dépannage

Erreurs courantes et correctifs

401 from Massive

Check API key. Verify it's still active in the dashboard.

DataFrame limit exceeded (50 / 50k rows default)

Raise via env config or chunk via SQL limits.

SQL syntax error in query_data

Query engine is a subset — check docs for supported functions.

Alternatives

mcp_massive vs autres

Alternative	Quand l'utiliser	Compromis
Alpaca MCP	You want trading too, not just data	Different broker; narrower data
Polygon.io / Alpha Vantage direct + fetch MCP	You want full API control	You write the fetching and parsing yourself

Plus

Ressources

📖 Lire le README officiel sur GitHub

🐙 Voir les issues ouvertes

🔍 Parcourir les 400+ serveurs MCP et Skills