csv-sift

juniobc/csv-sift

★ 0 stars Python Data/Analytics Updated 3d ago

MCP server to query a CSV that doesn't fit in context: profile, filter, aggregate, value-counts, search. pandas, read-only, no LLM.

View on GitHub → 🔍 Audit Wallet Slippage →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "csv-sift": {
      "command": "uvx",
      "args": [
        "csv-sift"
      ]
    }
  }
}

Or install with pip: pip install csv-sift

README Excerpt

An MCP server for the moment an assistant needs to work with a CSV it **can't fit in context**. Point it at a file; the model gets five tools that each return only the answer, not the file. No LLM, no network — just pandas behind a small, safe, whitelisted surface. | Tool | What it answers | |---|---|

Tools (5)

csv_aggregatecsv_profilecsv_querycsv_searchcsv_value_counts

📊

Working with market or time-series data?

The Polymarket Historical Dataset — 18.5M+ price snapshots across 18,600+ prediction markets (CSV + SQLite). Built for backtesting & ML.

View dataset →

csv-sift

Quick Install

README Excerpt

Tools (5)

Related Data/Analytics Servers