# shadcn/improve

Use your most capable model to audit your codebase and write plans for cheaper models to execute.

Repository: https://github.com/shadcn/improve
Canonical: https://ross.abutalabs.com/products/improve
License: MIT
License Family: permissive
Last push: 2026-06-15T06:18:39+00:00

## Health v2 (maintenance only)
Score: 53/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 87, release rhythm 35, longevity 6
- inputs: {"age_days": 84, "days_push": 79, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 8997, forks 404 (observed 2026-08-28T04:10:26.700349+00:00)

## What it is
An agent skill that audits a codebase using a capable LLM and writes self-contained markdown implementation plans for cheaper models to execute. It never implements changes itself; the plan is the product, with commands for audits, planning, execution dispatch, and backlog reconciliation.

## Use cases
- audit my codebase for tech debt and bugs
- generate implementation plans for AI coding agents
- use an expensive model to plan and a cheap model to implement
- review only what my current branch changes before a PR
- dispatch a cheaper model to execute a plan and review its diff
- keep a backlog of agent tasks reconciled across sessions
- get feature suggestions for where to take my project

## When to choose
- you use LLM coding agents and want to split planning from execution to cut costs
- you want structured, reviewable markdown specs instead of ad-hoc agent prompts
- you work in any agent that supports the Agent Skills format

## When to avoid
- you want a tool that automatically implements and merges fixes
- you don't use LLM-based coding agents in your workflow
- you need a standalone GUI or CI-integrated static analyzer rather than an agent skill

## Facets
- artifact type: plugin
- maturity: active
- function: developer-tools, code-review, workflow-automation, agent-framework, prompt-engineering
- domain: developer-tools, version-control
- platform: cli, cross-platform
- tags: agent-skills, code-audit, implementation-plans, llm-orchestration, code-quality, ai-agents, automation

## Member repositories
- shadcn/improve (main) score 53

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:26.700349+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:24:22.037518+00:00, confidence not recorded.
  - readme: https://github.com/shadcn/improve (fetched 2026-08-28T04:10:26.700349+00:00, sha 653a7a855936)
- Data as of 2026-08-30T08:39:29.467469+00:00.