Self-hosted, Self-Improving, Federated Everything App — dev management, agent orchestration, persistent mind, digital twin, second brain, and creative pipeline from a single dashboard over Tailscale to port the user into the OS
JavaScript ai-agentsdashboarddeveloper-toolsdigital-twinquantified-selfsecond-brainself-hostedtailscale
import { describe, it, expect, vi } from 'vitest';
import { render, screen, fireEvent } from '@testing-library/react';
import VisionDescribeModal from './VisionDescribeModal';
import { awaitEnabled } from '../../test/enabledBarrier.js';
// One enabled API provider so the action buttons are enabled-by-provider.
vi.mock('../../hooks/useProviderModels', () => ({
default: () => ({
providers: [{ id: 'ollama', name: 'Ollama', type: 'api', enabled: true }],
selectedProviderId: 'ollama',
selectedModel: 'qwen-vl',
availableModels: ['qwen-vl'],
setSelectedProviderId: () => {},
setSelectedModel: () => {},
loading: false,
}),
}));
// VisionProviderPicker unions the server's authoritative VLM set into its model
// filter, so it runs the capability scan on mount — stub it out (this suite is
// about the modal's actions, and the picker's own suite covers the union).
vi.mock('../../hooks/useVisionModelIds', () => ({
default: () => ({ idsByProvider: null, loaded: true }),
}));
// The gallery picker pulls in the media/socket layer — stub it.
vi.mock('../imageGen/GalleryImagePicker', () => ({ default: () => null }));
vi.mock('../ProviderModelSelector', () => ({ default: () => null }));
// Stub the upload helper so adding an image enables the actions without real I/O.
vi.mock('../../services/apiMedia', () => ({
processScreenshotUploads: vi.fn(async (_files, { onSuccess }) => {
onSuccess({ filename: 'up.png', preview: 'data:image/png;base64,x' });
}),
}));
const apiMocks = vi.hoisted(() => ({
describeEntityFromImages: vi.fn(),
expandEntityFromImages: vi.fn(),
}));
vi.mock('../../services/apiUniverseBuilder', () => apiMocks);
describe('VisionDescribeModal', () => {
const baseProps = {
open: true, entryName: 'Freydis', universeId: 'uni-1', entryId: 'chr-1',
onApply: () => {}, onApplyFields: () => {}, onClose: () => {},
};
it('shows the "Build character details" action for characters', () => {
render(<VisionDescribeModal {...baseProps} kind="character" />);
expect(screen.getByRole('button', { name: /Build character details/i })).toBeInTheDocument();
// Both image sources are offered.
expect(screen.getByRole('button', { name: /Gallery/i })).toBeInTheDocument();
});
it('hides the structured action for non-character kinds', () => {
render(<VisionDescribeModal {...baseProps} kind="place" />);
expect(screen.queryByRole('button', { name: /Build character details/i })).not.toBeInTheDocument();
// The prose describe action is still present for places.
expect(screen.getByRole('button', { name: /Describe from image/i })).toBeInTheDocument();
});
it('applies only the checked, edited attributes to onApplyFields', async () => {
const onApplyFields = vi.fn();
apiMocks.expandEntityFromImages.mockResolvedValue({
fields: { pronouns: 'she/her', age: 'late 20s' },
updatedFields: ['pronouns', 'age'],
llm: { provider: 'ollama', model: 'qwen-vl' },
});
render(<VisionDescribeModal {...baseProps} kind="character" onApplyFields={onApplyFields} />);
// Add an image so the actions enable (the file helper is stubbed). The modal
// portals to document.body, so query the document, not the render container.
fireEvent.change(document.querySelector('input[type="file"]'), {
target: { files: [new File(['x'], 'up.png', { type: 'image/png' })] },
});
const buildBtn = screen.getByRole('button', { name: /Build character details/i });
await awaitEnabled(buildBtn);
fireEvent.click(buildBtn);
// Review list renders both proposed fields; edit pronouns and uncheck age.
const pronouns = await screen.findByRole('textbox', { name: 'Pronouns' });
fireEvent.change(pronouns, { target: { value: 'she/they' } });
fireEvent.click(screen.getByRole('checkbox', { name: /Age/i }));
fireEvent.click(screen.getByRole('button', { name: /Apply 1 detail/i }));
// Edited value wins; unchecked field is dropped.
expect(onApplyFields).toHaveBeenCalledWith({ pronouns: 'she/they' });
});
it('keeps the description and attribute panels both open, closing only after the last is applied', async () => {
const onApply = vi.fn();
const onApplyFields = vi.fn();
const onClose = vi.fn();
apiMocks.describeEntityFromImages.mockResolvedValue({ description: 'a weathered barbarian', llm: {} });
apiMocks.expandEntityFromImages.mockResolvedValue({
fields: { pronouns: 'she/her' }, updatedFields: ['pronouns'], llm: {},
});
render(
<VisionDescribeModal {...baseProps} kind="character"
onApply={onApply} onApplyFields={onApplyFields} onClose={onClose} />,
);
fireEvent.change(document.querySelector('input[type="file"]'), {
target: { files: [new File(['x'], 'up.png', { type: 'image/png' })] },
});
// Describe → prose panel appears.
const describeBtn = screen.getByRole('button', { name: /Describe from image/i });
await awaitEnabled(describeBtn);
fireEvent.click(describeBtn);
await screen.findByRole('textbox', { name: /Generated description/i });
// Build → structured panel appears WITHOUT clearing the description.
fireEvent.click(screen.getByRole('button', { name: /Build character details/i }));
await screen.findByRole('textbox', { name: 'Pronouns' });
expect(screen.getByRole('textbox', { name: /Generated description/i })).toBeInTheDocument();
// Apply the attributes → modal stays open (description still unsaved).
fireEvent.click(screen.getByRole('button', { name: /Apply 1 detail/i }));
expect(onApplyFields).toHaveBeenCalled();
expect(onClose).not.toHaveBeenCalled();
// Apply the description → now the modal closes.
fireEvent.click(screen.getByRole('button', { name: /Apply description/i }));
expect(onApply).toHaveBeenCalledWith('a weathered barbarian');
expect(onClose).toHaveBeenCalled();
});
});