cbemister/claude-code-utils

accessibility-audit

Identify and fix accessibility issues for WCAG 2.1 Level AA compliance, including mobile screen reader support (VoiceOver/TalkBack), touch targets, and keyboard navigation.

原文语言：英语

build-app

Autonomously build an app from staged plans created by /launch-app. Reads plans/active/ stage files and builds each stage with verification, commits, and Slack notifications. Designed for unattended execution via build-app-runner.sh.

原文语言：英语

component-polish

Polish components to production quality — interactive states, micro-interactions, and final refinement. Run all three passes or target one area.

原文语言：英语

conversion-audit

Audit and optimize pages for conversion — value props, CTAs, copy, social proof, and friction points. Run a full audit or target a specific area.

原文语言：英语

design-app

Generate multiple design concepts for an app, compare them side-by-side, and implement the chosen direction. Creates detailed design documents with color palettes, typography, spacing, component specs, and interactive HTML mockups.

原文语言：英语

design-system

Apply foundational design system improvements — color palette, typography, spacing, and layout. Run all at once or target a specific area. Use before component-polish for a complete design overhaul.

原文语言：英语

enhance-app

Enhance an existing project with Claude Code agents, teams, rules, hooks, plans, MCP servers, and CLAUDE.md. Scans what exists, installs what's missing. Links shared files (agents, plan templates) from ~/.claude/shared/enterprise/ instead of duplicating.

原文语言：英语

enhance-design

Comprehensive design enhancement that chains all design skills in phases - analyze, mobile optimize, conversion optimize, and polish. Transforms any interface into a professional, high-converting, mobile-friendly experience.

原文语言：英语

evolution-gate

Use when an optimization has been built and needs preview deployment then human decision. Deploys preview, then approves (merge to production) or rejects (discard and retry).

原文语言：英语

factory

Use when managing the software factory lifecycle — launching new products, building them, evolving them for revenue, or checking factory-wide status.

原文语言：英语

generate-hypotheses

Use when evaluation data exists and optimization proposals are needed. Analyzes scores, user journey drop-off points, and prior attempts to produce ranked hypotheses batched 2-3 per cycle.

原文语言：英语

launch-app

Launch a new professional app from idea to staged build plan. Creates the project folder, configures the agent team, and generates all stage plans — ready for the team to build.

原文语言：英语

mobile-design

Apply mobile-first design patterns and touch interactions. Implements navigation, responsive layouts, touch targets, gestures, and platform-aware patterns for iOS and Android.

原文语言：英语

plan-status

Show current status and progress of active feature plans. Pass a stage filename to refresh it against the current codebase.

原文语言：英语

team-battle

Run two agent teams head-to-head on the same task. Creates parallel worktrees, installs different team agents in each, runs the task, and compares results side-by-side.

原文语言：英语

verify-work

Comprehensive pre-commit verification for security, best practices, code standards, and performance. Checks for vulnerabilities, efficiency issues, N+1 queries, and convention adherence.

原文语言：英语

worktree

Manage git worktrees for parallel feature development. Sub-commands: create, sync, cleanup

原文语言：英语

evaluate-product

Use when a built product needs scoring across conversion, revenue, UX, performance, accessibility, completeness, and code quality. Chains existing audit skills into a composite 0-100 score with AI-simulated user journeys.

原文语言：英语

plan-optimization

Use when hypotheses have been generated and need to be converted into an executable stage plan. Produces a standard stage plan that /build-app executes unchanged.

原文语言：英语

starter-project

Generate starter projects pre-configured with Claude Code agents and skills. Choose from SaaS app, API service, component library, CLI tool, e-commerce, or browser game templates.

原文语言：英语

ship

Complete end-of-session workflow - verify work quality, organize commits, and summarize what shipped. Use at the end of a development session to ship cleanly.

原文语言：英语

critique-value

Evaluate work from an end-user perspective — value delivered, usability, design quality, completeness, and accessibility. Scores each dimension and prioritizes improvements by user impact.

原文语言：英语

pm-review

Product manager perspective — assess current product state against the plan, verify what's actually built, identify gaps, and recommend what to build next ordered by user impact.

原文语言：英语

style

Transform UI into a chosen aesthetic. Supports 10 themes — aurora, brutalist, cyberpunk, dark-premium, glassmorphism, minimalist, neumorphism, organic, retro, swiss. Invoke as `/style <theme>` or without argument to choose interactively.

原文语言：英语

summarize-session

Capture a structured session work journal — what was built, decisions made, current state, and next steps. Writes to project memory for future session context.

原文语言：英语

browser-test

Run comprehensive Playwright browser tests across desktop, tablet, and mobile viewports. Auto-discovers pages and tests navigation, forms, auth, CRUD, responsive behavior, interactive elements, error states, user flows, and performance metrics.

原文语言：英语

create-plan

Initialize a feature development plan with phases and task tracking

原文语言：英语

generate-tests

Generate comprehensive test files for new or existing code. Creates test suites with proper mocking, edge cases, and error handling.

原文语言：英语

organize-commits

Organize staged changes into logical commits with conventional commit messages. Groups related changes and creates well-structured commit history.

原文语言：英语