More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Component Generator: React Components Worth Keeping

    1

    Generate React/TypeScript components that stay maintainable: typed props contract first, correct server/client boundary, behavior tests, and a story per state.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Component Generator: React Components Worth Keeping

    Component Generator: React Components Worth Keeping

    Example session with this skill installed

    Build me a PricingToggle component: it switches a pricing page between monthly and annual billing, shows the annual savings percentage, and reports the selected billing period to the parent. It must work in a React Server Components app.

    • Read your context and instructions
    • Compiled the component generator

    PricingToggle — generated with the skill, ready for review

    Props contract (written before any JSX):

    • period: "monthly" | "annual" — the selected billing period, required. A union, not two booleans, so the two states can never contradict each other.
    • annualSavingsPercent: number — required, e.g. 20; rendered as "Save 20% with annual".
    • onPeriodChange: (period) => void — the only callback; the parent owns the billing state.
    • defaultPeriod?: "monthly" | "annual" — defaults to "monthly".

    Boundary decision: the toggle is a small client leaf ('use client') — it owns only the pressed state and the click handler. The pricing cards stay server-rendered; no secrets or data fetching cross the boundary. Two real buttons with aria-pressed, so the selected state is announced and keyboard-accessible.

    Tests (behavior, not internals): renders both options with the savings label; clicking Annual calls onPeriodChange("annual"); clicking the already-selected option does not fire the callback again; the savings label is absent when annualSavingsPercent is 0; the default period renders selected on first paint.

    Stories: MonthlySelected, AnnualSelected, ZeroSavings, NarrowViewport (verifies the toggle still fits a mobile pricing page). Rubric score: 9/10 — one point held back because the savings label is hard-coded English; localization is a follow-up, not a blocker. No prop exists to work around a parent problem, and the contract fits in four props.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Most generated components rot because they are built backwards: JSX first, props discovered along the way, no tests, no stories. This skill reverses that order. It starts from a props contract — required vs optional decided deliberately, unions over booleans, every prop documented in one line — then places the server/client boundary so only the interactive leaves run on the client. The component itself does one job, takes data instead of services, and composes through children rather than prop sprawl.

    You get behavior tests that assert on what the user sees (not class names or hook calls), stories for every meaningful state so the component can be changed safely later, and naming rules that survive contact with a real team. A refactor decision table tells you exactly when to split a component, and a scoring rubric gates delivery: nothing ships below 8 out of 10. A worked example builds a searchable user list end to end — contract, split between server shell and client search leaf, tests, and stories — so the method is concrete, not abstract.

    • It cannot settle undecided design questions — if the interface is not decided, the generator will surface the gap rather than invent an answer.
    • It assumes React and TypeScript; the boundary guidance is specific to Server Components and does not map directly to other frameworks.
    • The rubric is a quality gate, not a guarantee: a component can score well and still be the wrong component for the product.
    • Styling guidance is structural (layout in the parent, appearance in the component), not a design system.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 3 days ago

    Permissions

    No special permissions declared or detected

    Listed3 days ago

    What's inside

    Frequently Asked Questions

    No. It generates individual components to a consistent quality bar. A design system is the tokens, patterns, and governance around those components — this skill makes the components worthy of one, it does not build the system itself.

    That is the split signal, not a rule to break quietly. The decision table in the skill shows the options: split into smaller components, use a compound component, or pass a cohesive data object the component actually owns. Ten props is the smell, not the crime.

    Because implementation changes and behavior should not. Tests that assert class names or hook call order break on harmless refactors and teach teams to delete tests. Behavior tests survive refactors and catch real regressions.