Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Build progressive forms with Conform. Use when creating forms that work without JavaScript, implementing server-validated forms in Remix/Next.js, or building accessible forms with progressive enhancement.
.claude/skills/terminalskills-conform/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | -44% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -28% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 8% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -27% | 0% |
Conform is a progressive enhancement form library. Forms work without JavaScript (server-side validation), then enhance with client-side validation when JS loads. Native to Remix and Next.js Server Actions. Uses Zod schemas for shared client/server validation.
typescript// app/actions.ts — Server action with Conform 'use server' import { parseWithZod } from '@conform-to/zod' import { z } from 'zod' const schema = z.object({ name: z.string().min(2), email: z.string().email(), message: z.string().min(10).max(1000), }) export async function submitContact(prevState: unknown, formData: FormData) { const submission = parseWithZod(formData, { schema }) if (submission.status !== 'success') { return submission.reply() // return errors to form } await db.contacts.create(submission.value) return submission.reply({ resetForm: true }) }
tsx// app/contact/page.tsx — Client form with progressive enhancement 'use client' import { useForm } from '@conform-to/react' import { parseWithZod } from '@conform-to/zod' import { useActionState } from 'react' import { submitContact } from '../actions' export default function ContactPage() { const [lastResult, action] = useActionState(submitContact, undefined) const [form, fields] = useForm({ lastResult, onValidate({ formData }) { return parseWithZod(formData, { schema }) // client-side validation }, shouldValidate: 'onBlur', shouldRevalidate: 'onInput', }) return ( <form id={form.id} onSubmit={form.onSubmit} action={action} noValidate> <div> <label htmlFor={fields.name.id}>Name</label> <input {...getInputProps(fields.name, { type: 'text' })} /> <p>{fields.name.errors}</p> </div> <div> <label htmlFor={fields.email.id}>Email</label> <input {...getInputProps(fields.email, { type: 'email' })} /> <p>{fields.email.errors}</p> </div> <div> <label htmlFor={fields.message.id}>Message</label> <textarea {...getTextareaProps(fields.message)} /> <p>{fields.message.errors}</p> </div> <button type="submit">Send</button> </form> ) }
shouldValidate: 'onBlur' validates when user leaves field — less aggressive than onChange.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 6,804 | 3,376 | -50% | 1 | 1 | 0% | 1,355 | 1,401 | +3% | 0 | 0 | — |
case-02 | pass→pass | 9,620 | 6,233 | -35% | 1 | 1 | 0% | 1,837 | 1,964 | +7% | 0 | 0 | — |
case-03 | fail→pass | 14,893 | 3,132 | -79% | 1 | 1 | 0% | 2,343 | 1,322 | -44% | 0 | 0 | — |
case-04 | pass→pass | 10,613 | 3,987 | -62% | 1 | 1 | 0% | 1,875 | 1,449 | -23% | 0 | 0 | — |
case-05 | fail→pass | 8,158 | 3,852 | -53% | 1 | 1 | 0% | 1,384 | 1,424 | +3% | 0 | 0 | — |
case-06 | pass→pass | 6,386 | 3,231 | -49% | 1 | 1 | 0% | 1,174 | 1,300 | +11% | 0 | 0 | — |
case-07 | pass→pass | 5,324 | 2,302 | -57% | 1 | 1 | 0% | 972 | 1,057 | +9% | 0 | 0 | — |
case-08 | pass→pass | 5,533 | 3,390 | -39% | 1 | 1 | 0% | 1,085 | 1,337 | +23% | 0 | 0 | — |
case-09 | pass→pass | 9,646 | 8,523 | -12% | 1 | 1 | 0% | 1,580 | 2,019 | +28% | 0 | 0 | — |
case-10 | pass→pass | 16,688 | 9,832 | -41% | 1 | 1 | 0% | 3,134 | 2,799 | -11% | 0 | 0 | — |
case-11 | fail→pass | 9,173 | 3,067 | -67% | 1 | 1 | 0% | 1,725 | 1,248 | -28% | 0 | 0 | — |
case-12 | pass→pass | 12,980 | 9,882 | -24% | 1 | 1 | 0% | 2,294 | 2,410 | +5% | 0 | 0 | — |
case-13 | fail→pass | 7,281 | 4,028 | -45% | 1 | 1 | 0% | 1,276 | 1,372 | +8% | 0 | 0 | — |
case-14 | fail→pass | 11,733 | 4,892 | -58% | 1 | 1 | 0% | 2,241 | 1,632 | -27% | 0 | 0 | — |
case-15 | fail→pass | 8,000 | 3,077 | -62% | 1 | 1 | 0% | 1,392 | 1,167 | -16% | 0 | 0 | — |
case-16 | pass→pass | 9,909 | 3,726 | -62% | 1 | 1 | 0% | 1,833 | 1,325 | -28% | 0 | 0 | — |
case-17 | fail→pass | 12,448 | 7,733 | -38% | 1 | 1 | 0% | 1,926 | 2,314 | +20% | 0 | 0 | — |
case-18 | pass→pass | 13,969 | 8,922 | -36% | 1 | 1 | 0% | 2,602 | 2,359 | -9% | 0 | 0 | — |
case-19 | pass→pass | 3,580 | 8,178 | +128% | 1 | 1 | 0% | 606 | 1,138 | +88% | 0 | 0 | — |
case-20 | pass→pass | 11,759 | 8,736 | -26% | 1 | 1 | 0% | 1,962 | 2,203 | +12% | 0 | 0 | — |
case-21 | pass→pass | 11,989 | 7,694 | -36% | 1 | 1 | 0% | 2,487 | 2,244 | -10% | 0 | 0 | — |
case-22 | pass→pass | 12,889 | 8,926 | -31% | 1 | 1 | 0% | 2,441 | 2,369 | -3% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +32 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.