Get Shit Done: The AI Coding System That Actually Works

Vibecoding has a reputation problem. You describe what you want, AI generates code, and you get inconsistent garbage that falls apart at scale.

It does not have to be this way.

Get Shit Done (GSD) is a free, open-source meta-prompting system that makes AI coding assistants like Claude Code reliable. It solves "context rot" — the quality degradation that happens as AI fills its context window.

And here is the kicker: it was built by a solo developer who does not write code anymore. Claude Code does.

What Is GSD?

GSD is a context engineering layer for AI coding assistants. It works with:

• Claude Code
• OpenCode
• Gemini CLI
• Codex
• GitHub Copilot
• Antigravity

The system handles the complexity so you do not have to. Under the hood, it uses XML prompt formatting, subagent orchestration, and state management. What you see: a few commands that just work.

How It Works

GSD gives Claude everything it needs to do the work and verify it. The workflow is simple:

1. Map Your Codebase

Already have code? Run /gsd:map-codebase first. It spawns parallel agents to analyze your stack, architecture, conventions, and concerns. Then /gsd:new-project knows your codebase — questions focus on what you are adding, and planning automatically loads your patterns.

2. Start a New Project

Run /gsd:new-project and the system:

• Questions — Asks until it understands your idea completely (goals, constraints, tech preferences, edge cases)
• Research — Spawns parallel agents to investigate the domain (optional but recommended)
• Requirements — Extracts what is v1, v2, and out of scope
• Roadmap — Creates phases mapped to requirements

You approve the roadmap. Then you are ready to build.

3. Discuss Each Phase

Run /gsd:discuss-phase 1 to shape the implementation. The system analyzes the phase and identifies gray areas based on what you are building:

• Visual features → Layout, density, interactions, empty states
• APIs/CLIs → Response format, flags, error handling, verbosity
• Content systems → Structure, tone, depth, flow

The deeper you go here, the more accurately the system builds.

Why It Matters for Indie Hackers

Most spec-driven development tools are enterprise overkill. They assume you are running a 50-person engineering org with sprint ceremonies, story points, stakeholder syncs, and Jira workflows.

If you are a solo builder, you do not need any of that. You need to describe what you want and have it built correctly.

GSD removes the enterprise theater while keeping the engineering rigor. It is:

• Lightweight — No complex setup, no ceremonies
• Powerful — Context engineering, XML formatting, subagent orchestration under the hood
• Free — Open-source, published on npm

"By far the most powerful addition to my Claude Code. Nothing over-engineered. Literally just gets shit done."

Trusted by engineers at Amazon, Google, Shopify, and Webflow.

How to Get Started

Install GSD with one command:

The installer prompts you to choose your runtime (Claude Code, OpenCode, Gemini, etc.) and location (global for all projects or local for current project only).

For non-interactive install (Docker, CI, scripts):

Then verify with:

• Claude Code / Gemini: /gsd:help
• OpenCode: /gsd-help
• Codex: $gsd-help

The Bottom Line

GSD solves the biggest problem with AI coding assistants: unreliability. It does not make them smarter. It gives them the context they need to stay focused and accurate throughout your project.

For indie hackers who want to move fast without sacrificing quality, this is the missing layer between "vibecoding" and "actually shipping."

The tool is free. The approach is proven. The only question is: what are you waiting for?