Get Shit Done: A meta-prompting, context engineering and spec-driven dev system

🎯 4 Dominant Themes from the HN thread

1️⃣ Massive AI‑generated code claims (but questioned)

“With GSD, I was able to write 250K lines of code in less than a month, without prior knowledge of claude.” — prakashrj

The discussion repeatedly returns to the astonishing LOC numbers people are reporting and whether such volume can truly be trusted or reviewed.

2️⃣ Structured planning frameworks (GSD / Superpowers) are praised for speed & clarity

“GSD consistently gets me 95 % of the way there on complex tasks. It’s a huge productivity boost over vanilla Claude Code.” — unst

Users highlight how tools that force a clear planning phase, spec files, and CI/CD pipelines let them move faster and stay organized.

3️⃣ Over‑engineering & token waste are heavily criticized

“Superpowers is a highly overengineered piece of software that unfortunately does not get shit done, burns limits and takes ages while doing so.” — yolonir

Several commenters argue the meta‑frameworks add unnecessary ceremony, consume excessive tokens, and often replace simpler, more direct prompting.

4️⃣ Verification, testing & security are the real bottlenecks > “Having a structure is really helpful – I’ve used some similar prompt scaffolds, and the difference is very noticeable.” — observationist

The consensus is that generating code is easy; the hard part is ensuring it works, stays secure, and can be reliably reviewed—something many feel current tools ignore.