claude developer-tools code-generation context-engineering

Get Shit Done: A Meta-Prompting, Context Engineering and Spec-Driven Dev System

GitHub - gsd-build/get-shit-done: A light-weight and powerful meta-prompting, context engineering and spec-driven development system for Claude Code by TÂCHES.

github.com

March 17, 2026

16 min read

🔥🔥🔥🔥🔥

67/100

Summary

Get Shit Done is a lightweight meta-prompting and context engineering system designed for Claude Code, OpenCode, Gemini CLI, Codex, Copilot, and Antigravity. It addresses context rot and is compatible with Mac, Windows, and Linux, allowing users to efficiently build specifications without unnecessary complications.

Key Takeaways

GSD is a meta-prompting and context engineering system designed for Claude Code and other AI coding tools, aimed at improving development efficiency.
The system addresses context rot, enhancing the reliability of AI-generated code by providing necessary context and verification.
GSD is compatible with multiple platforms including Mac, Windows, and Linux, and is trusted by engineers at major companies like Amazon and Google.
Users can install GSD using the command `npx get-shit-done-cc@latest`, with options for global or local installation across various AI runtimes.

Read original article

View story →

Coverage & discussions

Hacker News · github.com

Get Shit Done: A Meta-Prompting, Context Engineering and Spec-Driven Dev System ↗

Community discussions

Hacker News discussion ↗

Historical count: 5 · 5 saved in sample · Historical sample

Analysis ready

Community Sentiment

Mixed

Positives

Using openspec has significantly boosted my productivity, allowing me to generate nearly all my code with high efficiency and deploy multiple times a day.
The comparison between frameworks like superpowers and this new system could provide valuable insights into optimizing AI-driven development workflows.

Concerns

The frameworks can consume up to 10 times more tokens, leading to concerns about cost-effectiveness and diminishing returns on outcomes.
One user found the output to be excessively verbose, generating too many files and complicating updates as user requirements evolved.