WhySpec - Open-source, spec-driven development framework for AI agents

Hey Product Hunt! 👋 I’m Quan, solo-founder of GitWhy.

Today we’re launching our first open-source project: WhySpec.

We built WhySpec because most AI development workflows today — even strong ones like OpenSpec, GSD, and gstack — still don’t preserve the Decision Bridge. They help you plan, execute, and debug, but the reasoning behind the current session usually disappears once the coding is done.

So we built a spec-driven context engineering framework that preserves that "why" from coding agents for future chats, future debugging, and future reviews.

WhySpec is the spec-driven development framework with reasoning layer for AI coding agents: an open-source CLI and agent skill set that helps you plan before coding, capture what actually happened after coding, and track the delta between the two. We call that the Decision Bridge — the gap between the decisions you expected to make and the ones you actually made.

It’s MIT licensed, works offline, needs no account, and is built to work across tools like Claude Code, Cursor, GitHub Copilot, Codex CLI, Windsurf, and more.

My goal with WhySpec is simple: make the reasoning behind code visible, searchable, and durable — especially in a world where agents write more and more of the implementation.

If this resonates, I’d really love your feedback, questions, and tough critiques in the comments. Thanks for checking out WhySpec 💛

Get started: npm install -g @gitwhy-cli/whyspec