chore: import upstream snapshot with attribution
CI / Shell Format Check (push) Has been cancelled
CI / Check Ruby (3.4) (push) Has been cancelled
CI / CI Config (push) Has been cancelled
CI / Test on Node ${{ matrix.node }} and ${{ matrix.os }}${{ matrix.shard && format(' (shard {0}/3)', matrix.shard) || '' }} (push) Has been cancelled
CI / Build on Node ${{ matrix.node }} (push) Has been cancelled
CI / Style Check (push) Has been cancelled
CI / Generate Assets (push) Has been cancelled
CI / Check Python (3.14) (push) Has been cancelled
CI / Check Python (3.9) (push) Has been cancelled
CI / Build Docs (push) Has been cancelled
CI / Code Scan Action (push) Has been cancelled
CI / Site tests (push) Has been cancelled
CI / webui tests (push) Has been cancelled
CI / Run Integration Tests (push) Has been cancelled
CI / Run Smoke Tests (push) Has been cancelled
CI / Go Tests (push) Has been cancelled
CI / Share Test (push) Has been cancelled
CI / Redteam (Production API) (push) Has been cancelled
CI / Redteam (Staging API) (push) Has been cancelled
CI / GitHub Actions Lint (push) Has been cancelled
CI / Check Ruby (3.0) (push) Has been cancelled
release-please / release-please (push) Has been cancelled
release-please / build (push) Has been cancelled
release-please / publish-npm (push) Has been cancelled
release-please / publish-npm-backfill (push) Has been cancelled
release-please / docker (push) Has been cancelled
release-please / publish-code-scan-action (push) Has been cancelled
release-please / attest-code-scan-action (push) Has been cancelled
Deploy local.promptfoo.app / Deploy to Cloudflare Pages (push) Has been cancelled
Test and Publish Multi-arch Docker Image / test (push) Has been cancelled
Test and Publish Multi-arch Docker Image / build-docker-and-push-digests (map[digest-suffix:linux-amd64 platform:linux/amd64 runner:ubuntu-latest]) (push) Has been cancelled
Test and Publish Multi-arch Docker Image / build-docker-and-push-digests (map[digest-suffix:linux-arm64 platform:linux/arm64 runner:ubuntu-24.04-arm]) (push) Has been cancelled
Test and Publish Multi-arch Docker Image / merge-docker-digests (push) Has been cancelled
Test and Publish Multi-arch Docker Image / Attest Multi-arch Image (push) Has been cancelled
Validate Renovate Config / Validate Renovate Configuration (push) Has been cancelled
CI / Shell Format Check (push) Has been cancelled
CI / Check Ruby (3.4) (push) Has been cancelled
CI / CI Config (push) Has been cancelled
CI / Test on Node ${{ matrix.node }} and ${{ matrix.os }}${{ matrix.shard && format(' (shard {0}/3)', matrix.shard) || '' }} (push) Has been cancelled
CI / Build on Node ${{ matrix.node }} (push) Has been cancelled
CI / Style Check (push) Has been cancelled
CI / Generate Assets (push) Has been cancelled
CI / Check Python (3.14) (push) Has been cancelled
CI / Check Python (3.9) (push) Has been cancelled
CI / Build Docs (push) Has been cancelled
CI / Code Scan Action (push) Has been cancelled
CI / Site tests (push) Has been cancelled
CI / webui tests (push) Has been cancelled
CI / Run Integration Tests (push) Has been cancelled
CI / Run Smoke Tests (push) Has been cancelled
CI / Go Tests (push) Has been cancelled
CI / Share Test (push) Has been cancelled
CI / Redteam (Production API) (push) Has been cancelled
CI / Redteam (Staging API) (push) Has been cancelled
CI / GitHub Actions Lint (push) Has been cancelled
CI / Check Ruby (3.0) (push) Has been cancelled
release-please / release-please (push) Has been cancelled
release-please / build (push) Has been cancelled
release-please / publish-npm (push) Has been cancelled
release-please / publish-npm-backfill (push) Has been cancelled
release-please / docker (push) Has been cancelled
release-please / publish-code-scan-action (push) Has been cancelled
release-please / attest-code-scan-action (push) Has been cancelled
Deploy local.promptfoo.app / Deploy to Cloudflare Pages (push) Has been cancelled
Test and Publish Multi-arch Docker Image / test (push) Has been cancelled
Test and Publish Multi-arch Docker Image / build-docker-and-push-digests (map[digest-suffix:linux-amd64 platform:linux/amd64 runner:ubuntu-latest]) (push) Has been cancelled
Test and Publish Multi-arch Docker Image / build-docker-and-push-digests (map[digest-suffix:linux-arm64 platform:linux/arm64 runner:ubuntu-24.04-arm]) (push) Has been cancelled
Test and Publish Multi-arch Docker Image / merge-docker-digests (push) Has been cancelled
Test and Publish Multi-arch Docker Image / Attest Multi-arch Image (push) Has been cancelled
Validate Renovate Config / Validate Renovate Configuration (push) Has been cancelled
This commit is contained in:
@@ -0,0 +1,173 @@
|
||||
import { beforeEach, describe, expect, it, vi } from 'vitest';
|
||||
import { fetchWithCache } from '../../../src/cache';
|
||||
import { BeavertailsPlugin } from '../../../src/redteam/plugins/beavertails';
|
||||
import { CustomPlugin } from '../../../src/redteam/plugins/custom';
|
||||
import { CyberSecEvalPlugin } from '../../../src/redteam/plugins/cyberseceval';
|
||||
import { DoNotAnswerPlugin } from '../../../src/redteam/plugins/donotanswer';
|
||||
import { HarmbenchPlugin } from '../../../src/redteam/plugins/harmbench';
|
||||
import { Plugins } from '../../../src/redteam/plugins/index';
|
||||
import { IntentPlugin } from '../../../src/redteam/plugins/intent';
|
||||
import { PlinyPlugin } from '../../../src/redteam/plugins/pliny';
|
||||
import { UnsafeBenchPlugin } from '../../../src/redteam/plugins/unsafebench';
|
||||
import { shouldGenerateRemote } from '../../../src/redteam/remoteGeneration';
|
||||
import {
|
||||
createMockProvider,
|
||||
createProviderResponse,
|
||||
type MockApiProvider,
|
||||
} from '../../factories/provider';
|
||||
|
||||
vi.mock('../../../src/cache');
|
||||
vi.mock('../../../src/cliState', () => ({
|
||||
__esModule: true,
|
||||
default: { remote: false },
|
||||
}));
|
||||
vi.mock('../../../src/redteam/remoteGeneration', async (importOriginal) => {
|
||||
return {
|
||||
...(await importOriginal()),
|
||||
getRemoteGenerationUrl: vi.fn().mockReturnValue('http://test-url'),
|
||||
neverGenerateRemote: vi.fn().mockReturnValue(false),
|
||||
shouldGenerateRemote: vi.fn().mockReturnValue(false),
|
||||
};
|
||||
});
|
||||
vi.mock('../../../src/util', async () => ({
|
||||
...(await vi.importActual('../../../src/util')),
|
||||
|
||||
maybeLoadFromExternalFile: vi.fn().mockReturnValue({
|
||||
generator: 'Generate test prompts',
|
||||
grader: 'Grade the response',
|
||||
}),
|
||||
}));
|
||||
vi.mock('../../../src/integrations/huggingfaceDatasets', async (importOriginal) => {
|
||||
return {
|
||||
...(await importOriginal()),
|
||||
|
||||
fetchHuggingFaceDataset: vi
|
||||
.fn()
|
||||
.mockResolvedValue([{ vars: { prompt: 'test prompt', category: 'test' } }]),
|
||||
};
|
||||
});
|
||||
|
||||
// Mock contracts plugin to ensure it has canGenerateRemote = true
|
||||
vi.mock('../../../src/redteam/plugins/contracts', async () => {
|
||||
const original = await vi.importActual('../../../src/redteam/plugins/contracts');
|
||||
return {
|
||||
...original,
|
||||
};
|
||||
});
|
||||
|
||||
describe('canGenerateRemote property and behavior', () => {
|
||||
let mockProvider: MockApiProvider;
|
||||
|
||||
beforeEach(() => {
|
||||
mockProvider = createMockProvider({
|
||||
response: createProviderResponse({
|
||||
output: 'Sample output',
|
||||
error: null as any,
|
||||
}),
|
||||
});
|
||||
|
||||
// Reset all mocks
|
||||
vi.clearAllMocks();
|
||||
vi.mocked(fetchWithCache).mockReset();
|
||||
});
|
||||
|
||||
describe('Plugin canGenerateRemote property', () => {
|
||||
it('should mark dataset-based plugins as not requiring remote generation', () => {
|
||||
expect(BeavertailsPlugin.canGenerateRemote).toBe(false);
|
||||
expect(CustomPlugin.canGenerateRemote).toBe(false);
|
||||
expect(CyberSecEvalPlugin.canGenerateRemote).toBe(false);
|
||||
expect(DoNotAnswerPlugin.canGenerateRemote).toBe(false);
|
||||
expect(HarmbenchPlugin.canGenerateRemote).toBe(false);
|
||||
expect(IntentPlugin.canGenerateRemote).toBe(false);
|
||||
expect(PlinyPlugin.canGenerateRemote).toBe(false);
|
||||
expect(UnsafeBenchPlugin.canGenerateRemote).toBe(false);
|
||||
});
|
||||
});
|
||||
|
||||
describe('Remote generation behavior', () => {
|
||||
it('should not use remote generation for dataset-based plugins even when shouldGenerateRemote is true', async () => {
|
||||
vi.mocked(shouldGenerateRemote).mockImplementation(function () {
|
||||
return true;
|
||||
});
|
||||
|
||||
const unsafeBenchPlugin = Plugins.find((p) => p.key === 'unsafebench');
|
||||
|
||||
await unsafeBenchPlugin?.action({
|
||||
provider: mockProvider,
|
||||
purpose: 'test',
|
||||
injectVar: 'testVar',
|
||||
n: 1,
|
||||
config: {},
|
||||
delayMs: 0,
|
||||
});
|
||||
|
||||
expect(fetchWithCache).not.toHaveBeenCalled();
|
||||
});
|
||||
|
||||
it('should use remote generation for LLM-based plugins when shouldGenerateRemote is true', async () => {
|
||||
vi.mocked(shouldGenerateRemote).mockImplementation(function () {
|
||||
return true;
|
||||
});
|
||||
|
||||
// Force the canGenerateRemote property to be true for this test
|
||||
const originalContractPlugin = Plugins.find((p) => p.key === 'contracts');
|
||||
if (!originalContractPlugin) {
|
||||
throw new Error('Contract plugin not found');
|
||||
}
|
||||
|
||||
// Create a mock plugin with canGenerateRemote=true
|
||||
const mockContractPlugin = {
|
||||
...originalContractPlugin,
|
||||
action: vi.fn().mockImplementation(async function () {
|
||||
await fetchWithCache('http://test-url/api/generate', {
|
||||
method: 'POST',
|
||||
headers: { 'Content-Type': 'application/json' },
|
||||
body: JSON.stringify({ test: true }),
|
||||
});
|
||||
return [];
|
||||
}),
|
||||
};
|
||||
|
||||
// Call the mocked action
|
||||
await mockContractPlugin.action({
|
||||
provider: mockProvider,
|
||||
purpose: 'test',
|
||||
injectVar: 'testVar',
|
||||
n: 1,
|
||||
config: {},
|
||||
delayMs: 0,
|
||||
});
|
||||
|
||||
// Verify fetchWithCache was called
|
||||
expect(fetchWithCache).toHaveBeenCalledWith('http://test-url/api/generate', {
|
||||
method: 'POST',
|
||||
headers: { 'Content-Type': 'application/json' },
|
||||
body: expect.any(String),
|
||||
});
|
||||
});
|
||||
|
||||
it('should use local generation for all plugins when shouldGenerateRemote is false', async () => {
|
||||
vi.mocked(shouldGenerateRemote).mockImplementation(function () {
|
||||
return false;
|
||||
});
|
||||
|
||||
// Use the plugin from Plugins array directly
|
||||
const contractPlugin = Plugins.find((p) => p.key === 'contracts');
|
||||
if (!contractPlugin) {
|
||||
throw new Error('Contract plugin not found');
|
||||
}
|
||||
|
||||
await contractPlugin.action({
|
||||
provider: mockProvider,
|
||||
purpose: 'test',
|
||||
injectVar: 'testVar',
|
||||
n: 1,
|
||||
config: {},
|
||||
delayMs: 0,
|
||||
});
|
||||
|
||||
expect(fetchWithCache).not.toHaveBeenCalled();
|
||||
expect(mockProvider.callApi).toHaveBeenCalledWith(expect.any(String));
|
||||
});
|
||||
});
|
||||
});
|
||||
Reference in New Issue
Block a user