Scottcjn

Scottcjn / trashclaw

Public

Local LLM agent for vintage Mac hardware. Built on a 2013 Mac Pro trashcan with dual AMD FirePro D500s. Elyan Labs.

rustchain.org
13
40
100% credibility
Found Mar 23, 2026 at 13 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

TrashClaw is a local AI agent that uses built-in tools to perform file operations, shell commands, git tasks, and more on your computer based on natural language instructions.

How It Works

1
📖 Discover TrashClaw

You hear about TrashClaw, a friendly local helper that tackles computer tasks like organizing files and running commands on your machine.

2
⬇️ Get it ready

Download the single main file and example helpers to a folder on your computer—no extra installs needed.

3
🚀 Wake it up

Start TrashClaw and link it to your local AI thinker so it can understand and act on your ideas.

4
📁 Pick your project

Open TrashClaw in your work folder, and it shows your current spot with a welcoming prompt.

5
💭 Tell it a task

Type something simple like 'find all notes to fix and make a list,' and feel the excitement as it springs into action.

6
🔄 Watch the magic

TrashClaw reads files, thinks step-by-step, asks for okay on changes, and updates your project safely.

🎉 Task complete

Enjoy your finished work, like a new organized list saved and ready, with fun badges for your progress.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 13 to 13 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 trashclaw?

TrashClaw is a Python-based local LLM agent that turns natural language tasks into actions on your machine, like finding TODOs, creating git commits, or editing files via any local LLM server such as Ollama or llama.cpp. It solves the need for an offline, zero-dependency coding assistant that handles file ops, shell commands, git workflows, and web fetches without cloud APIs. Point it at a local LLM for coding on vintage hardware like a 2013 Mac Pro, and it iterates with tools until the job's done.

Why is it gaining traction?

It stands out with single-file setup, no pip installs, and a plugin system for custom tools, making it a lightweight local GitHub Copilot alternative for repo tasks or a local GitHub actions runner without servers. Devs dig the git-aware prompts, undo for file changes, tab completion, and stats like tokens/sec, plus it benchmarks local LLMs on obscure hardware. The hook: drop-in plugins and commands like /status or /diff keep sessions productive across projects.

Who should use this?

Backend devs managing local repositories who want a local LLM for coding tasks like git commits or code stats without internet. Tinkerers on 2013-era Mac hardware or vintage setups needing a local LLM agent for file diffs and shell automation. Solo hackers avoiding cloud costs for quick local GitHub runner-style workflows.

Verdict

Worth a spin for local LLM enthusiasts on old iron—docs are solid, features pack a punch despite 13 stars and 1.0% credibility score signaling early maturity. Expect occasional multi-step flubs with small models; pair with a beefier local LLM server for reliability.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.