fltman

fltman / cosmo

Public

AI-powered data analytics with GPU-accelerated graph visualization. Two AI agents debate your data autonomously.

11
1
100% credibility
Found Apr 03, 2026 at 11 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Cosmo is an AI analytics tool where users drop CSV or JSON files and two AI agents autonomously analyze, debate, generate charts, and manipulate an interactive graph visualization.

How It Works

1
🔍 Discover Cosmo

You stumble upon Cosmo, a friendly tool that makes sense of your spreadsheets and data files with smart helpers.

2
🚀 Set it up quickly

Download the files and connect a thinking service so the AIs can start helping you right away.

3
📁 Drop in your data

Drag any CSV or JSON file onto the screen, and it loads up instantly for exploration.

4
💬 Start chatting

Type something simple like 'show me insights' and watch the first AI analyze your data on the spot.

5
🤝 Bring in a partner

Flip the switch to enable a second AI that debates and digs deeper into the findings.

6
📊 Watch it come alive

Beautiful charts pop up, the graph zooms to important spots, and connections light up before your eyes.

🎉 Unlock hidden stories

You gain clear, actionable understanding of your data, like a personal expert team working for you.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 11 to 11 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 cosmo?

Cosmo is an AI-powered data analytics tool built in Python with a FastAPI backend and Vite frontend. Drop CSV or JSON files (even multiples) into the web app at localhost:5199, ask natural questions like "give me insights," and two agents—Claude as analyst, Gemini as companion—autonomously debate findings, generate charts, and manipulate a GPU-accelerated graph visualization powered by Cosmograph. It handles multi-dataset joins, dynamic relationships, and reusable plugins, turning raw data into interactive explorations without writing code.

Why is it gaining traction?

The killer hook is the real-time multi-agent debate: enable the companion in modes like Devil's Advocate or Anomaly Hunter, and watch AIs challenge each other while updating charts (bar, scatter, pie via Chart.js) and the live graph (zoom, filter, recolor). Unlike basic AI-powered data viz tools, it executes custom Pandas ops on-the-fly, detects cross-dataset links, and offers full transparency on every tool call. Devs dig the autonomous depth—no more solo LLM hallucinations.

Who should use this?

Data analysts prototyping insights on sales CSVs, devops teams scanning codebases via the filesystem tool for bloat or patterns, or indie devs visualizing user metrics in JSON. Perfect for anyone tired of manual Pandas graphing who wants AI-powered data management and visualization in a single drop-and-chat interface.

Verdict

Early alpha (11 stars, 1.0% credibility score) with excellent README quickstart but no tests or production hardening—fork and tweak for your ai-powered projects github stack. Try it if agent-driven graph debates spark joy; skip for mission-critical pipelines.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.