mnemosyne-systems

mnemosyne-systems / orangu

Public

Advanced code editor

ai-coding cli code-editor coding-assistant developer-tools
10
3
85% credibility
Found May 18, 2026 at 10 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

orangu is a terminal-based coding assistant that connects to local AI models to help you work on your code. It can read files in your project, make edits, run shell commands, and fetch web content—all while you chat with it naturally. The AI uses built-in 'tools' to take actions in your workspace, making it feel like pair programming with an intelligent assistant that respects your project's boundaries and keeps everything private on your machine.

How It Works

1
💡 You want an AI coding buddy that lives on your computer

You've heard about AI assistants but want one that runs locally, keeps your code private, and actually understands your project files.

2
📦 You download and build orangu

The tool is free and open source. You grab the code and build it on your machine in just a few commands.

3
⚙️ You point it at your AI server and project folder

You write a simple config file telling orangu where your AI model is running and which folder contains your code.

4
🚀 You launch orangu and start chatting

A colorful terminal window opens showing your workspace status. You type natural questions like 'show me the main file' or 'help me fix this bug'.

5
The AI gets to work on your request
📖
It reads your files

You ask to see a file, and orangu instantly shows it with nice formatting and line numbers.

✏️
It edits your code

You ask it to make a change, and it updates your files directly—no copy-pasting required.

🔧
It runs commands

You ask it to test something, and it runs shell commands in your project and shows you the results.

6
💬 You keep chatting until it's done

The conversation flows naturally. You can ask follow-up questions, cancel requests with Escape, or queue up commands while waiting for responses.

✅ Your code is fixed, your file is updated, your question is answered

You've got a helpful AI assistant that works entirely on your machine, respects your project boundaries, and never shares your code with anyone.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 10 to 10 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 orangu?

orangu is a terminal-based coding assistant built in Rust that connects to OpenAI-compatible servers, with special support for llama.cpp. It gives you a persistent chat interface where an AI can actually read, edit, and explore your local project files. The tool runs locally, stays within your workspace boundaries, and lets you pipe natural language commands like "open README.md" or "list files" directly into the session. Think of it as a smarter shell that understands your codebase.

Why is it gaining traction?

The main draw is that this isn't another cloud-hosted AI wrapper. orangu keeps everything on your machine, which matters for privacy and latency. Its tight integration with llama.cpp means you get native streaming metrics (tokens per second, prompt progress) in the footer while you wait. The workspace-aware tools mean the model can actually modify files, run shell commands, and browse your project tree without hallucinating paths. It's a refreshing contrast to services that just return text.

Who should use this?

Developers running local LLMs who want an interactive coding partner rather than a one-shot prompt box. If you're experimenting with llama.cpp on a beefy workstation and want something between a raw API call and a full IDE plugin, this fills that gap. Not for teams needing enterprise support or Windows users (tested on Fedora only).

Verdict

At version 0.3.0 with only 10 stars, orangu is clearly early-stage. The documentation exists and the feature set is coherent, but it's a project for enthusiasts rather than production deployments. With a credibility score of 0.8500000238418579%, the fundamentals are solid -- the code is clean, the dependencies are reasonable, and the architecture makes sense. Worth watching if you want to run local AI coding tools without cloud dependencies.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.