Baraakhaldi

Feature Benchmark — Testing skill for Claude Code

Testing community

Claude Code skill: benchmark one feature of your app or website against the products that do it best, driven by your needs.

How to install Feature Benchmark

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Baraakhaldi/feature-benchmark and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Feature Benchmark does

Claude Code skill: benchmark one feature of your app or website against the products that do it best, driven by your needs. Outputs a cited brief, report with flow screenshots, feature matrix, spec and build prompt.

Alternatives in Testing

Playwright — claude-plugins-official Browser automation, E2E testing, screenshots 29.4k ★
Brainstorm Experiments Existing — Design experiments to test assumptions for existing products 7.8k ★
Flow Spec — NLSpec authoring — use when you need a structured specification from multi-AI research and consensus 2.8k ★

README

feature-benchmark

A [Claude Code](https://claude.com/claude-code) skill that benchmarks **one feature** of your app, website or product idea (search, onboarding, checkout, booking, filters, notifications, dashboard, chat, settings, sharing, anything) against the products that do it best, starting from **your needs**.

It is the feature-level sibling of [website-benchmark](https://github.com/Baraakhaldi/website-benchmark): instead of auditing a whole company website, you say which feature and what it must do, and everything (which products to compare, what to measure, how to score) follows from that.

What you get

Feature brief: ranked needs, best practices, anti-patterns, pattern choices, what users say, the standards and rules that apply, and a needs checklist. Every point cited.
Feature benchmark report: your feature scored against 5–15+ products on 8 feature dimensions, a feature matrix (✓ ◐ ✗ ?), the same flow step side by side across products with screenshots, issues as Problem → Screenshot → Who does it better → Research reason → Fix, the 10 best patterns and a roadmap.
Feature spec: user stories, target flow, every state with its copy, acceptance criteria, edge cases, accessibility, analytics events.
Component kit and resource directory: verified components, libraries and services to build it, with licences.
Build prompt: a ready-to-paste prompt for another Claude session to build or fix the feature.

How it works

Intake: the feature, your needs, and the product in any form (URL, app, screenshots, Figma, repo, or "not built yet"), plus extras like analytics, tickets or reviews.
Questions: multiple choice, pre-filled from feature presets: scope, status, users, ranked needs, success signal, platforms, where to look, depth, outputs, permissions.
Scope check, twice: a proposal you can edit, then Start · Edit again · Cancel.
Research: feature best practices, user expectations, releva