Radar / AI coding tools / shadcn/improve
shadcn/improve: audit with your best model, execute with your cheapest
An MIT-licensed Agent Skill that audits any codebase across nine categories with parallel subagents, vets its own findings, and writes self-contained implementation plans (with verification gates and stop conditions) for cheaper models to execute in isolated git worktrees.
Why it matters
The plan is the product: the skill never writes code itself, so teams pay frontier-model prices once for judgment and hand execution to cheap models — a cost-split workflow that hit #4 on GitHub daily trending on Nov 11 with 10,970 stars.
What you could build with it
A small team maintaining a large legacy codebase could run /improve weekly to turn a frontier model's audit into a prioritized backlog of self-contained plans, then let cheaper models work through the plans in disposable worktrees while humans only review the diffs.
Does it hold up?
Usable today and independently verified in a hands-on run: the tutorial's real audit produced evidence-backed plans with file-and-line citations, though the advisor run itself still burns frontier-model tokens and plan quality depends on the executor model chosen.
Built with shadcn/improve
- Hands-on with shadcn's improve: audit your codebase with your best model, execute with your cheapestaifrontierpost · Full hands-on tutorial run on Nov 11 on a demo repo with real planted issues, showing the actual plans produced; notes it ranked #4 on GitHub trending that day.
- shadcn/improve | There's An AI For Thattheresanaiforthat · Repo mirror with the full README: nine audit categories, /improve execute flow, and the hard rules (never modifies source, executors only in worktrees).
- Artificial Analysis' Coding Agent Index v2.0mindpattern.ai · Oct 24 roundup flagging improve as the skill that doesn't write code — its README promise that the plan is the product.
- shadcn/improvegithub · 10,971 ★ · The skill itself: SKILL.md, references, and an example plan, installable via `npx skills add shadcn/improve`.
Learn more
First spotted on github: source.
More AI coding tools
Chrome DevTools MCP: give your coding agent a real browserOfficial MCP server from the Chrome team that lets coding agents drive a live Chrome — console, network, traces…coding · JEV 0.82Orca: the agent development environment for running coding-agent fleetsStably AI's open-source desktop app for running many CLI coding agents side by side — Claude Code, Codex, Cursor and…coding · JEV 0.78Open Code Review: Alibaba's hybrid AI code reviewer, now past 50,000 starsAlibaba's open-source code review CLI (Apache 2.0) pairs deterministic rules with an LLM agent to review git diffs…coding · JEV 0.77Astryx: Meta's agent-ready design system, open sourceMeta open-sourced the React 19 + StyleX design system behind 13,000 internal apps: 150+ accessible components, themes…coding · JEV 0.76GitHub Spec Kit — spec-driven development for coding agentsGitHub's open-source toolkit that gives any coding agent a structured spec, plan, tasks, implement workflow, so specs…coding · JEV 0.75Superpowers: agentic skills framework that makes coding agents plan, test, and review before shippingobra/superpowers installs composable skills into Claude Code, Cursor, Codex and eight more harnesses, enforcing a…coding · JEV 0.74
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.