app-evaluationlisted
Install: claude install-skill cportka/claude-plugins
# App / Website Evaluation
Audit an app or website and hand back a **prioritized, evidence-backed** report: what's good,
what's missing, and the highest-impact fixes — ranked by impact × effort, each tied to concrete
evidence (a missing tag, a failing header, an absent file), never vague advice.
**Be self-referential.** There is no universal "good site." Judge each property against what's
best **for its type and its community**: a marketing site, a SaaS app, a docs site, a blog, an
open-source project, an e-commerce store, a portfolio, an API, or a mobile-app landing page each
have different priorities, standards, and growth channels. So **classify first**, then evaluate
against that class's norms. Apply the same lens to your own recommendations — the *best* PR and
software-design move for a niche dev tool is not the same as for a consumer app.
## 1. Classify the target (do this first)
Determine, from the URL/repo/screens and by asking if unclear:
- **Type** (marketing / SaaS / docs / blog / OSS project / e-commerce / portfolio / API / app
landing) and **primary goal** (signups, sales, stars, leads, reads, installs).
- **Audience & community** (developers, designers, consumers, a specific niche) — this drives
which channels, tone, and standards matter.
- **Stage** (pre-launch, launched, scaling) and **stack** (static, SSR, SPA, CMS) — a SPA has
different crawlability/perf concerns than a static site.
Everything below is weighted by these answers. State the classifi