Repo Source Code Test Frameworks logo

Repo Source Code Test Frameworks

OrganizationPopular
open-circle
repo-source-code-test-frameworks

Write unit and type tests for Formisch framework packages (frameworks/angular, frameworks/preact, frameworks/solid, frameworks/svelte, frameworks/vue, frameworks/react, frameworks/react-native). Use when adding tests for hooks/composables/runes/inject functions (useForm/createForm/injectForm, useField/injectField, useFieldArray/injectFieldArray) or components and directives (Form, Field, FieldArray, formischForm, formischField) in any framework wrapper.

Overview

Publisheropen-circle
Repositoryformisch
Skill namerepo-source-code-test-frameworks
Stars
1.2K
Forks
36
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by open-circle on GitHub. Read the source before you install it.

Installation

Install the Repo Source Code Test Frameworks AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/open-circle/formisch/tree/main/.agents/skills/repo-source-code-test-frameworks
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/open-circle/formisch.git /tmp/formisch
mkdir -p .claude/skills
cp -r /tmp/formisch/.agents/skills/repo-source-code-test-frameworks .claude/skills/repo-source-code-test-frameworks
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Repo Source Code Test Frameworks in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Repo Source Code Test Frameworks on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Repo Source Code Test Frameworks is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Writing Framework Tests

Tests for frameworks/<fw>/. For packages/core/ or packages/methods/, use repo-source-code-test-packages.

Rules

  1. React is the canonical reference. New shared tests land in frameworks/react first, then port.
  2. Cross-framework consistency. Shared API (useForm/createForm/injectForm, useField, useFieldArray, Field, FieldArray) keeps the same describe/test names, schemas, initialInput, expected outputs. Only mechanical adapters differ. Test what the adapter actually exports: React Native has no Form component, and Angular ships directives instead of components.
  3. No framework hint in test names. Never 'should ... (Solid-only)'. Use a // ... comment above the test if framework-specific reasoning matters.
  4. No tautological assertions. A test must be able to fail under a real source change.
  5. 100% coverage on non-excluded files. Match React's vitest.config.ts exclusions: src/types, src/vitest, **/index.ts(x), test files.
  6. vi.waitFor for async, never setTimeout(0).

File layout

frameworks/<fw>/
├── vitest.config.ts          # framework plugin + jsdom + setupFiles + coverage exclusions
├── package.json              # `test: vitest run --typecheck`, framework testing-library devDep
└── src/
    ├── vitest/
    │   ├── setup.ts          # jest-dom + cleanup
    │   ├── renderHook.ts     # (Vue/Svelte only)
    │   └── *Host.svelte      # (Svelte only) shared hosts
    ├── hooks|primitives|composables|runes|functions/<name>/<name>.test.{ts,tsx} + .test-d.ts
    └── components|directives/<Name>/<Name>.test.{ts,tsx}
                            └── *Host.test.svelte # (Svelte only) per-test hosts

Standard test set

Identical describe/test names across all frameworks:

Sourcedescribe blocksTests
useForm / createForm / injectForminitialization, initial validation, reactivity4 + 2 type
useField / injectFieldinitialization, input updates, validation modes, element registration10 + 4 type
useFieldArray / injectFieldArrayinitialization, reactivity6 + 4 type
Form / Field / FieldArray(flat)3 each

Form only exists where the adapter provides it: React Native has no Form component, and Angular ships the [formischForm] directive instead.

The store stability describe tests useMemo across re-renders and exists only in React and React Native. In other frameworks omit it with a one-line // Note: ... comment.

Per-framework adapters

ConcernPreactSolidSvelteVueReactReact NativeAngular
Hook nameuseFormcreateFormcreateFormuseFormuseFormuseForminjectForm
Read stateform.isValid.valueform.isValidform.isValidform.isValidform.isValidform.isValidform.isValid()
Imperative setfield.onInput(v)field.onInput(v)field.onInput(v)field.input = vfield.onChange(v)field.onChange(v)field.setInput(v)
DOM input eventfireEvent.inputfireEvent.inputfireEvent.inputinput.setValue(v)fireEvent.changefireEvent.changeinput.dispatchEvent(new Event())
Auto-focus propautofocusautofocusautofocusautofocusautoFocusnonenone
Field childrenfunctionfunction{#snippet children}v-slot / scoped slotfunctionfunction*formischField + let field
Testing library@testing-library/preact@solidjs/testing-library@testing-library/svelte@vue/test-utils@testing-library/react@testing-library/react@angular/core/testing (TestBed)

renderHook

  • Preact / React / React Native: from the testing library — { result: { current }, rerender, unmount }.
  • Solid: from @solidjs/testing-library — { result, owner, cleanup }. result is direct, not wrapped in .current. Use createRoot from solid-js only if no JSX context is needed; our primitives need an owner so renderHook is the default.
  • Vue: custom helper at src/vitest/renderHook.ts — mounts a defineComponent({ setup() { return hook(); } }) (the canonical Vue Test Utils TestComponent recipe). Required because onBeforeMount/computed need component context.
  • Angular: no renderHook. Call the inject functions inside TestBed.runInInjectionContext(() => …) (or runInInjectionContext(injector, …) when you need to control destruction), and mount a host @Component with TestBed.createComponent for directive tests.
  • Svelte: custom helper at src/vitest/renderHook.ts — mounts Hook.svelte. Required because onMount and [createAttachmentKey()] need template context. For pure rune logic with no lifecycle, $effect.root(() => { ... })() in a .test.svelte.ts is lighter — but no covered rune in this repo qualifies.

Async

ts
// ❌ flaky
await new Promise((r) => setTimeout(r, 0));
// ✅
await vi.waitFor(() => expect(form.isValid).toBe(false));

For Svelte rune updates, flushSync() from svelte synchronously drains derived/effect updates before the assertion. For Vue, flushPromises() from @vue/test-utils is fine for one-shot microtask drains; prefer vi.waitFor for multi-tick reactive flows.

Type tests (*.test-d.ts)

ts
import * as v from 'valibot';
import { describe, expectTypeOf, test } from 'vitest';
import type { FieldStore } from '../../types/index.ts';
import { useForm } from '../useForm/index.ts';
import { useField } from './useField.ts';

describe('useField', () => {
  test('should narrow input type for primitive leaves', () => {
    const schema = v.object({ name: v.string() });
    const form = useForm({ schema });
    expectTypeOf(useField(form, { path: ['name'] }).input).toEqualTypeOf<
      string | undefined
    >();
  });

  test('should reject invalid paths', () => {
    const form = useForm({ schema: v.object({ name: v.string() }) });
    // @ts-expect-error nonexistent field
    useField(form, { path: ['nonexistent'] });
  });
});

For Preact, wrap leaf types in ReadonlySignal<...>, and for Angular in its Signal<...> — those are the only framework-specific deltas in .test-d.ts files. Angular's types/field.test-d.ts additionally checks that a FieldStore stays assignable across schema and path generics, which is what its directives rely on.

Schema types

Always derive with typeof schema. If the schema is referenced only at the type level (host owns runtime), suppress the lint warning rather than hand-writing the valibot generic shape:

ts
// eslint-disable-next-line @typescript-eslint/no-unused-vars -- referenced only via `typeof schema`
const schema = v.object({ name: v.string() });
type Schema = typeof schema;

Framework gotchas

Angular — every reactive field.* / form.* is an Angular Signal<T>; read by calling it (field.isDirty()). Setter is field.setInput(v). There is no props object: templates bind [formischControl]="field" and pass the value with [value]="field.input() ?? ''". Field tests mount a host @Component whose template uses *formischField="['path'] of form; let field", configure TestBed with provideZonelessChangeDetection(), and drive the DOM with input.dispatchEvent(new Event('input')) followed by await fixture.whenStable() and fixture.detectChanges(). src/vitest/setup.ts already calls TestBed.initTestEnvironment and resets the module after each test.

Preact — every reactive field.* / form.* is a ReadonlySignal<T>; read with .value. Setter is field.onInput(v). DOM events fire on oninput. props.autofocus is lowercase.

React Native — tests render through react-native-web with @testing-library/react and jsdom, so fireEvent.change / fireEvent.focus work on the rendered TextInput. field.props has no name and no autoFocus; text is set via onChangeText. There is no Form component and no submit method, so submission tests call handleSubmit. Element cleanup differs by React version: react-native-web discards ref cleanup functions, so tests that assert per-element unregistration must render raw elements — see the comments in useField.test.tsx.

Solid — primitive is createForm (not useForm). Methods (insert, swap) don't need act(). Use class= not className= in JSX. vite-plugin-solid ≥ 2.8.2 auto-configures vitest; don't add manual deps overrides.

Vue — field.props has no onInput; wire DOM input via the setter:

ts
h('input', {
  ...field.props,
  value: field.input ?? '',
  onInput: (e: Event) => {
    field.input = (e.target as HTMLInputElement).value;
  },
});

For "element removed" assertions, mount with attachTo: document.body and query document.querySelector — wrapper.find keeps finding stale nodes after unmount().

Svelte — runes need Svelte compilation. Three test patterns:

  • Hook.svelte via renderHook — default; required when source calls onMount or uses [createAttachmentKey()].
  • *Host.test.svelte per-test components — for snippet/slot rendering of Field/FieldArray/Form.
  • $effect.root in a *.test.svelte.ts — only for pure rune logic with no lifecycle.

Snippets can't run statements; forward values via {@const _ = onField(field)} with a comment explaining the pattern. After every test, cleanup() from @testing-library/svelte must run in setup.ts or screen.getByTestId finds duplicates.

Svelte build hygiene — package.json files excludes **/*.test.*, **/*.test-d.*, dist/vitest. Verify after adding new fixtures:

bash
pnpm -C frameworks/svelte build && cd frameworks/svelte && pnpm pack --pack-destination /tmp/
tar -tzf /tmp/formisch-svelte-*.tgz | grep -E "(test|Host|vitest)"  # must be empty

Coverage

bash
cd frameworks/<fw> && pnpm exec vitest run --coverage

Must hit 100% on every non-excluded file. Common gaps that need framework-specific tests (no React analogue): field.props.onChange handler (Preact/Solid/Svelte/Vue, covered by validate:'change' test), usePathSignal length-mismatch branch (Preact), unwrap (Solid/Svelte). Add such tests with a // ... comment, not a (X-only) suffix.

Running

bash
pnpm -C frameworks/<fw> test           # runtime + type
pnpm -C frameworks/<fw> lint           # eslint + tsc --noEmit

CI: <fw>_vitest job in .github/workflows/ci.yml.

Checklist

  • Test names match React canonical names exactly (no (X-only) suffixes)
  • Framework divergence justified by // ... comment, not test name
  • vi.waitFor for async, never setTimeout(0)
  • Schema types via typeof schema, never hand-written valibot generics
  • No tautological assertions
  • 100% coverage on covered files
  • pnpm test and pnpm lint pass
  • (Svelte) pnpm pack output contains no test/Host/vitest files
  • CI has a <fw>_vitest job

Frequently asked questions

What does the Repo Source Code Test Frameworks AI skill do?

Write unit and type tests for Formisch framework packages (frameworks/angular, frameworks/preact, frameworks/solid, frameworks/svelte, frameworks/vue, frameworks/react, frameworks/react-native). Use when adding tests for hooks/composables/runes/inject functions (useForm/createForm/injectForm, useField/injectField, useFieldArray/injectFieldArray) or components and directives (Form, Field, FieldArray, formischForm, formischField) in any framework wrapper.

Why use Repo Source Code Test Frameworks on TypingMind?

Because you install it once and use it with any model. Repo Source Code Test Frameworks is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Repo Source Code Test Frameworks in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/open-circle/formisch/tree/main/.agents/skills/repo-source-code-test-frameworks. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Repo Source Code Test Frameworks?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Repo Source Code Test Frameworks?

As many as you like. As long as a model supports skills, you can use Repo Source Code Test Frameworks with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Repo Source Code Test Frameworks AI skill free?

Yes. It is published on GitHub by open-circle under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇