Relentlessly interviews you about a plan or design until every decision branch is resolved.
How I did it
AI-assisted, human-directed
Turns the current conversation context into a structured PRD and submits it as a GitHub issue.
Breaks a plan, spec, or PRD into independently-grabbable GitHub issues using vertical slices.
Explores a codebase to find opportunities for architectural improvement, focusing on reuse, coupling, and clarity.
Reviews recently changed code for reuse, quality, and efficiency, then fixes any issues found.
Audits copy for patterns common in AI-generated text, such as inflated vocabulary, filler phrases, and em dash overuse, then rewrites it.
Writes a conventional commit message from the staged changes.
Tech stack
Tools I build with
Anthropic's agentic CLI. Runs skills, manages context, and executes multi-step engineering tasks inside your terminal.
All issues, PRs, and code review happen here. The structured issue workflow keeps work visible and reviewable.
Hosts and deploys the site. Pushing to main deploys it, so there is no infrastructure to manage.
Handles DNS and proxies traffic in front of Vercel. Adds caching, DDoS protection, and keeps the domain config in one place.
App Router for routing and layout; Tailwind for styling. Both have broad enough community support that Claude Code reasons about them reliably.