mbzuai-oryx

mbzuai-oryx / OpenEarthAgent

Public

OpenEarthAgent is a unified framework for tool-augmented geospatial agents.

mbzuai-oryx.github.ioOpenEarthAgent
29
1
100% credibility
Found Feb 24, 2026 at 17 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

OpenEarthAgent is a framework that lets AI agents analyze satellite imagery and maps using specialized tools for detection, measurements, and change analysis.

How It Works

1
🌍 Discover OpenEarthAgent

You find this helpful tool for exploring satellite images and earth data on its project page.

2
📥 Download and prepare

Grab the program files and set up the pieces needed to run it on your computer.

3
🚀 Start the smart helpers

Turn on the background services that provide analysis tools like measurements and detections.

4
💬 Open the chat window

Launch the friendly web chat or simple command line to begin your earth exploration.

5
🖼️ Upload your image

Share a satellite photo or map and ask a question about features, changes, or measurements.

6
🔍 See step-by-step analysis

Watch the agent use tools to detect objects, compute distances, or analyze vegetation automatically.

✅ Get clear insights

Receive easy-to-understand answers, annotated maps, and data summaries about your earth image.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 17 to 29 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 OpenEarthAgent?

OpenEarthAgent is a Python framework for building tool-augmented geospatial agents that reason over satellite imagery and GIS data. It handles multi-step tasks like urban analysis, environmental monitoring, and disaster response by chaining tools for object detection, spectral indices (NDVI, NBR), POI mapping, and distance calculations. Users get CLI chat, Gradio web demos, pre-trained models on Hugging Face, plus training and evaluation scripts on a 14k-instance dataset.

Why is it gaining traction?

It unifies perceptual, GIS, and spectral tools under a single agent interface, skipping the hassle of stitching separate libraries. Developers notice the interpretable reasoning traces and live tool execution in demos, with caching for efficiency. The arXiv-backed dataset and HF models lower the entry barrier for custom fine-tuning.

Who should use this?

Remote sensing engineers analyzing multispectral imagery for change detection or infrastructure assessment. ML researchers prototyping geospatial agents for disaster response or urban planning. GIS analysts needing natural-language queries over satellite data without manual scripting.

Verdict

Early but solid starter for tool-augmented geospatial agents—try the Gradio demo if you're in remote sensing. With just 17 stars and 1.0% credibility, it's immature (complex multi-env setup, limited tests), so prototype lightly until docs stabilize.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.