FEAT: GUI view existing scores #2189
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,42 @@ | ||
| import { makeStyles, tokens } from '@fluentui/react-components' | ||
|
|
||
| export const useAttackVerdictChipStyles = makeStyles({ | ||
| chip: { | ||
| display: 'inline-flex', | ||
| alignItems: 'center', | ||
| columnGap: tokens.spacingHorizontalXS, | ||
| }, | ||
| chipLabel: { | ||
| textTransform: 'capitalize', | ||
| }, | ||
| surface: { | ||
| display: 'flex', | ||
| flexDirection: 'column', | ||
| rowGap: tokens.spacingVerticalXS, | ||
| minWidth: '240px', | ||
| maxWidth: '360px', | ||
| }, | ||
| row: { | ||
| display: 'flex', | ||
| columnGap: tokens.spacingHorizontalS, | ||
| }, | ||
| rowLabel: { | ||
| minWidth: '72px', | ||
| color: tokens.colorNeutralForeground2, | ||
| }, | ||
| rationaleBlock: { | ||
| display: 'flex', | ||
| flexDirection: 'column', | ||
| rowGap: tokens.spacingVerticalXXS, | ||
| marginTop: tokens.spacingVerticalXS, | ||
| paddingTop: tokens.spacingVerticalXS, | ||
| borderTop: `1px solid ${tokens.colorNeutralStroke2}`, | ||
| }, | ||
| rationaleText: { | ||
| color: tokens.colorNeutralForeground2, | ||
| whiteSpace: 'pre-wrap', | ||
| wordBreak: 'break-word', | ||
| maxHeight: '30vh', | ||
| overflowY: 'auto', | ||
| }, | ||
| }) |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,62 @@ | ||
| import { render, screen, within } from '@testing-library/react' | ||
| import userEvent from '@testing-library/user-event' | ||
| import { FluentProvider, webLightTheme } from '@fluentui/react-components' | ||
| import AttackVerdictChip from './AttackVerdictChip' | ||
| import type { ScoreView } from '../../types' | ||
|
|
||
| const TestWrapper: React.FC<{ children: React.ReactNode }> = ({ children }) => ( | ||
| <FluentProvider theme={webLightTheme}>{children}</FluentProvider> | ||
| ) | ||
|
|
||
| const sampleScore: ScoreView = { | ||
| id: 'score-1', | ||
| scorer_type: 'SelfAskRefusalScorer', | ||
| score_type: 'true_false', | ||
| score_value: 'true', | ||
| score_category: ['refusal'], | ||
| score_rationale: 'The model refused the request.', | ||
| timestamp: '2026-01-15T11:00:00Z', | ||
| } | ||
|
|
||
| describe('AttackVerdictChip', () => { | ||
| it('renders nothing when there is no score', () => { | ||
| render( | ||
| <TestWrapper> | ||
| <AttackVerdictChip outcome="failure" score={null} /> | ||
| </TestWrapper> | ||
| ) | ||
|
|
||
| expect(screen.queryByTestId('attack-score-chip')).not.toBeInTheDocument() | ||
| }) | ||
|
|
||
| it('shows only the score value on the chip, labeled "Score"', () => { | ||
| render( | ||
| <TestWrapper> | ||
| <AttackVerdictChip outcome="failure" score={sampleScore} /> | ||
| </TestWrapper> | ||
| ) | ||
|
|
||
| const chip = screen.getByRole('button', { name: /score true/i }) | ||
| expect(within(chip).getByText('Score')).toBeInTheDocument() | ||
| expect(within(chip).getByText('true')).toBeInTheDocument() | ||
| // The outcome word must never appear on the chip itself. | ||
| expect(within(chip).queryByText(/failure/i)).not.toBeInTheDocument() | ||
| }) | ||
|
|
||
| it('opens a popover with full score details when clicked', async () => { | ||
| const user = userEvent.setup() | ||
| render( | ||
| <TestWrapper> | ||
| <AttackVerdictChip outcome="failure" score={sampleScore} /> | ||
| </TestWrapper> | ||
| ) | ||
|
|
||
| await user.click(screen.getByRole('button', { name: /score true/i })) | ||
|
|
||
| const details = await screen.findByTestId('attack-score-details') | ||
| expect(within(details).getByText('true_false')).toBeInTheDocument() | ||
| expect(within(details).getByText('SelfAskRefusalScorer')).toBeInTheDocument() | ||
| expect(within(details).getByText('refusal')).toBeInTheDocument() | ||
| expect(within(details).getByText('The model refused the request.')).toBeInTheDocument() | ||
| }) | ||
| }) |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,86 @@ | ||
| import { | ||
| Button, | ||
| Text, | ||
| Badge, | ||
| Popover, | ||
| PopoverTrigger, | ||
| PopoverSurface, | ||
| } from '@fluentui/react-components' | ||
| import type { AttackOutcome, ScoreView } from '../../types' | ||
| import { resolveOutcome } from '../../utils/attackOutcome' | ||
| import { getScoreColor } from '../../utils/scoreColor' | ||
| import { useAttackVerdictChipStyles } from './AttackVerdictChip.styles' | ||
|
|
||
| interface AttackVerdictChipProps { | ||
| outcome?: AttackOutcome | null | ||
| score?: ScoreView | null | ||
| } | ||
|
|
||
| export default function AttackVerdictChip({ outcome, score }: AttackVerdictChipProps) { | ||
| const styles = useAttackVerdictChipStyles() | ||
|
|
||
| // Only the score value is surfaced in the ribbon, so there's nothing to show without a score. | ||
| if (!score) return null | ||
|
Contributor
There was a problem hiding this comment. Early-returning on
Contributor
There was a problem hiding this comment. I agree, maybe we can show at least the attack outcome badge (exactly as is shown in the attack table), and if scores are present who them in the click/hover effect? so attack verdict means (outcome + score), and the AttackVerdictChip shows exactly those |
||
|
|
||
| const resolvedOutcome = resolveOutcome(outcome) | ||
| const categories = score.score_category?.filter(Boolean) ?? [] | ||
| const scoreColor = getScoreColor(score.score_type, score.score_value) | ||
| const scoreBadgeStyle = { | ||
| backgroundColor: scoreColor.background, | ||
| borderColor: scoreColor.background, | ||
| color: scoreColor.foreground, | ||
| } | ||
|
|
||
| return ( | ||
| <Popover withArrow> | ||
| <PopoverTrigger disableButtonEnhancement> | ||
| <Button | ||
| appearance="subtle" | ||
| className={styles.chip} | ||
| data-testid="attack-score-chip" | ||
| aria-label={`Score ${score.score_value}`} | ||
| > | ||
| <span className={styles.chipLabel}>Score</span> | ||
| <Badge appearance="filled" size="medium" style={scoreBadgeStyle}> | ||
| {score.score_value} | ||
| </Badge> | ||
| </Button> | ||
| </PopoverTrigger> | ||
| <PopoverSurface> | ||
| <div className={styles.surface} data-testid="attack-score-details"> | ||
| <Text weight="semibold" size={300}>Score</Text> | ||
| <div className={styles.row}> | ||
| <Text size={200} weight="semibold" className={styles.rowLabel}>Value</Text> | ||
| <Badge appearance="filled" size="small" style={scoreBadgeStyle}> | ||
| {score.score_value} | ||
| </Badge> | ||
| </div> | ||
| <div className={styles.row}> | ||
| <Text size={200} weight="semibold" className={styles.rowLabel}>Type</Text> | ||
| <Text size={200}>{score.score_type}</Text> | ||
| </div> | ||
| <div className={styles.row}> | ||
| <Text size={200} weight="semibold" className={styles.rowLabel}>Scorer</Text> | ||
| <Text size={200}>{score.scorer_type}</Text> | ||
| </div> | ||
| {categories.length > 0 && ( | ||
| <div className={styles.row}> | ||
| <Text size={200} weight="semibold" className={styles.rowLabel}>Category</Text> | ||
| <Text size={200}>{categories.join(', ')}</Text> | ||
| </div> | ||
| )} | ||
| <div className={styles.row}> | ||
| <Text size={200} weight="semibold" className={styles.rowLabel}>Outcome</Text> | ||
| <Text size={200} className={styles.chipLabel}>{resolvedOutcome}</Text> | ||
| </div> | ||
| {score.score_rationale && ( | ||
| <div className={styles.rationaleBlock}> | ||
| <Text size={200} weight="semibold">Rationale</Text> | ||
| <Text size={200} className={styles.rationaleText}>{score.score_rationale}</Text> | ||
| </div> | ||
| )} | ||
| </div> | ||
| </PopoverSurface> | ||
| </Popover> | ||
| ) | ||
| } | ||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,41 @@ | ||
| import { makeStyles, tokens } from '@fluentui/react-components' | ||
|
|
||
| export const useObjectiveHeaderStyles = makeStyles({ | ||
| root: { | ||
| flexShrink: 0, | ||
| display: 'flex', | ||
| flexDirection: 'row', | ||
| alignItems: 'baseline', | ||
| columnGap: tokens.spacingHorizontalS, | ||
| padding: `${tokens.spacingVerticalS} ${tokens.spacingHorizontalL}`, | ||
| backgroundColor: tokens.colorNeutralBackground2, | ||
| borderBottom: `1px solid ${tokens.colorNeutralStroke1}`, | ||
| borderLeft: `3px solid ${tokens.colorBrandStroke1}`, | ||
| }, | ||
| label: { | ||
| flexShrink: 0, | ||
| }, | ||
| content: { | ||
| flexGrow: 1, | ||
| minWidth: 0, | ||
| color: tokens.colorNeutralForeground1, | ||
| fontSize: tokens.fontSizeBase300, | ||
| }, | ||
| contentCollapsed: { | ||
| whiteSpace: 'nowrap', | ||
| overflow: 'hidden', | ||
| textOverflow: 'ellipsis', | ||
| }, | ||
| contentExpanded: { | ||
| whiteSpace: 'pre-wrap', | ||
| wordBreak: 'break-word', | ||
| maxHeight: '30vh', | ||
| overflowY: 'auto', | ||
| }, | ||
| toggle: { | ||
| flexShrink: 0, | ||
| minWidth: 'auto', | ||
| whiteSpace: 'nowrap', | ||
| color: tokens.colorBrandForeground1, | ||
| }, | ||
| }) |
Oops, something went wrong.
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Minor: every other branch here defaults
objectiveto'', but the success path assignsattack.objectivedirectly. It's typed as required so this is fine today — worth a?? ''for consistency/robustness if the backend ever omits it.