gobbleyourdong

gobbleyourdong / tsunami

Public

autonomous AI agent. local models. no cloud. no API keys.

12
3
100% credibility
Found Apr 03, 2026 at 12 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Tsunami is a local AI agent that autonomously builds and tests software projects from natural language instructions using open-source models on your hardware.

How It Works

1
🔍 Discover Tsunami

You hear about Tsunami, a friendly helper that builds apps and tools just by telling it what you want, all on your own computer.

2
📥 Get it running

Download the simple installer or run one easy command, and it sets everything up automatically, finding your computer's power.

3
🚀 Start your helper

Open Tsunami, and it wakes up ready, downloading what it needs and opening a chat window.

4
💭 Tell it what to make

Type something fun like 'build me a calculator' or 'make a pinball game', and watch it spring into action.

5
👀 Watch it create

See your helper research ideas, write code, test everything, and fix issues until it's perfect.

✅ Enjoy your new creation

Your app appears in a web preview, fully working and ready to use or share with friends.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 12 to 12 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 tsunami?

Tsunami is a local autonomous AI agent framework built in Python that turns natural language prompts into working apps—no cloud, no API keys, just your hardware. Tell it "build me a calculator" or "replicate Game Boy UI," and it researches, generates images with SD-Turbo, grounds UI elements via vision, writes React code, compiles with Vite, tests interactions, and iterates until it passes QA. It auto-detects GPU/RAM to run Qwen 9B/2B LLMs in full or lite mode, delivering projects to a web UI at localhost:9876.

Why is it gaining traction?

Unlike cloud-bound autonomous agents like GitHub Copilot or Microsoft tools, tsunami runs everything offline with multi-agent capabilities— a central planner dispatches parallel workers for tasks like file analysis or shell execution. One-click installs via curl/setup.sh/Docker handle models and deps, and it includes React scaffolds for dashboards, data viz, and forms. Developers dig the tension system that detects hallucinations and forces verification, plus desktop/CLI UIs for real-time monitoring.

Who should use this?

Frontend devs prototyping React apps or UIs from sketches, indie hackers building MVPs without subscription costs, or AI tinkerers exploring autonomous agents in software development on local hardware. Ideal for quick experiments like analyzing file batches or generating grounded web components, not production pipelines.

Verdict

Try it if you're into github autonomous coding agents and have 8GB+ GPU—impressive local autonomy despite 12 stars and 1.0% credibility score signaling early-stage flux. Polish docs and add tests before relying on it daily, but it's a solid autonomous agents github find for offline prototyping.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.