jgravelle

jgravelle / jdatamunch-mcp

Public

Token-efficient MCP server for tabular data retrieval. Index CSV/Excel files, query rows, aggregate — 99%+ token savings vs raw file reads.

14
1
100% credibility
Found Mar 29, 2026 at 14 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

jDataMunch-MCP indexes local CSV and Excel files into a structured format for AI agents to retrieve dataset descriptions, filtered rows, and aggregations with drastically reduced token usage.

How It Works

1
📊 Discover smarter data analysis

You have a huge spreadsheet and want your AI helper to explore it without wasting money on reading every single row.

2
🔧 Set up the organizer

You add this simple organizer tool to your AI setup in just a few moments.

3
📁 Prepare your spreadsheet

You tell the organizer about your CSV or Excel file, and it neatly organizes the data once for quick access.

4
✨ Unlock precise insights

Your AI now gets column summaries, filtered rows, and smart calculations instantly, skipping irrelevant data.

5
🔍 Explore with ease

Ask your AI to describe the data, search columns, grab specific rows, or summarize groups effortlessly.

🎉 Save time and money

You analyze massive spreadsheets efficiently, seeing huge reductions in costs while getting spot-on answers.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 14 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is jdatamunch-mcp?

jDataMunch-MCP is a Python MCP server that indexes CSV/Excel files into a local SQLite store for token-efficient data retrieval. It solves the problem of AI agents dumping entire raw files into prompts, wasting tokens on irrelevant rows—instead, agents can query filtered rows, aggregate data server-side, and get schema profiles with 99%+ savings versus raw reads. Install via pip, index files with `index_local`, then use MCP tools like `describe_dataset` or `get_rows`.

Why is it gaining traction?

It delivers precise retrieval—search columns, filter with operators like `eq` or `between`, run GROUP BY aggregates—all without re-reading files, plus telemetry tracks your exact token and cost savings. Local-first with incremental indexing beats cloud data tools for speed and privacy, and MCP compatibility plugs right into Claude or similar agents. Benchmarks on 1M-row CSVs show 25,000x reductions, hooking devs optimizing LLM budgets.

Who should use this?

AI agent builders analyzing sales CSVs or crime stats, finance analysts querying Excel reports via LLMs, or ops engineers filtering large data files without full reloads. Ideal for anyone whose prompts bloat with unused tabular rows during exploration or aggregation.

Verdict

Worth a spin for MCP users with CSV/Excel workflows—proven 99%+ savings make it a smart retrieval layer despite 14 stars and 1.0% credibility score. Early-stage with strong docs and benchmarks, but verify indexing speed on your files first.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.