Add unit tests for skill capabilities, skill planner, and UV setup
- Implement tests for random ID generation, ensuring preference for crypto.randomUUID. - Create tests for runtime context capabilities, validating the injection of enabled skill capabilities. - Add tests for skill capability parsing, including classification and command example extraction. - Introduce tests for the skill planner, verifying tool call planning based on user requests and attachment requirements. - Establish tests for UV setup, ensuring proper handling of Python installation scenarios and environment checks.
This commit is contained in:
DEV_DSW
@@ -4,10 +4,18 @@ import { beforeEach, describe, expect, it, vi } from 'vitest';
|
||||
|
||||
const mocks = vi.hoisted(() => {
|
||||
const sessionMessages: any[] = [];
|
||||
const activeRuns = new Map<string, { runId: string; abortController: AbortController }>();
|
||||
|
||||
return {
|
||||
sessionMessages,
|
||||
activeRuns,
|
||||
providerChat: vi.fn(),
|
||||
providerGetCapabilities: vi.fn(),
|
||||
getEnabledSkillCapabilities: vi.fn(() => []),
|
||||
toolRuntimeRun: vi.fn(),
|
||||
createChatToolRuntime: vi.fn(() => ({
|
||||
run: mocks.toolRuntimeRun,
|
||||
})),
|
||||
appendMessage: vi.fn((_: string, message: unknown) => {
|
||||
sessionMessages.push(message);
|
||||
}),
|
||||
@@ -16,8 +24,13 @@ const mocks = vi.hoisted(() => {
|
||||
messages: [...sessionMessages],
|
||||
updatedAt: Date.now(),
|
||||
})),
|
||||
setActiveRun: vi.fn(),
|
||||
clearActiveRun: vi.fn(),
|
||||
setActiveRun: vi.fn((sessionKey: string, runId: string, abortController: AbortController) => {
|
||||
activeRuns.set(sessionKey, { runId, abortController });
|
||||
}),
|
||||
clearActiveRun: vi.fn((sessionKey: string) => {
|
||||
activeRuns.delete(sessionKey);
|
||||
}),
|
||||
getActiveRun: vi.fn((sessionKey: string) => activeRuns.get(sessionKey)),
|
||||
appendTranscriptLine: vi.fn(),
|
||||
maybeHandleBrowserOpenMessage: vi.fn(() => false),
|
||||
maybeHandleSkillInstallMessage: vi.fn(() => false),
|
||||
@@ -30,6 +43,7 @@ const mocks = vi.hoisted(() => {
|
||||
vi.mock('@electron/providers', () => ({
|
||||
createProvider: vi.fn(() => ({
|
||||
chat: mocks.providerChat,
|
||||
getCapabilities: mocks.providerGetCapabilities,
|
||||
})),
|
||||
}));
|
||||
|
||||
@@ -53,6 +67,7 @@ vi.mock('../electron/gateway/session-store', () => ({
|
||||
getOrCreate: mocks.getOrCreate,
|
||||
setActiveRun: mocks.setActiveRun,
|
||||
clearActiveRun: mocks.clearActiveRun,
|
||||
getActiveRun: mocks.getActiveRun,
|
||||
},
|
||||
}));
|
||||
|
||||
@@ -60,6 +75,18 @@ vi.mock('@electron/utils/token-usage-writer', () => ({
|
||||
appendTranscriptLine: mocks.appendTranscriptLine,
|
||||
}));
|
||||
|
||||
vi.mock('../electron/gateway/skill-capability-registry', () => ({
|
||||
getEnabledSkillCapabilities: mocks.getEnabledSkillCapabilities,
|
||||
}));
|
||||
|
||||
vi.mock('../electron/gateway/chat-tooling', async () => {
|
||||
const actual = await vi.importActual<typeof import('../electron/gateway/chat-tooling')>('../electron/gateway/chat-tooling');
|
||||
return {
|
||||
...actual,
|
||||
createChatToolRuntime: mocks.createChatToolRuntime,
|
||||
};
|
||||
});
|
||||
|
||||
vi.mock('../electron/gateway/browser-shortcut', () => ({
|
||||
maybeHandleBrowserOpenMessage: mocks.maybeHandleBrowserOpenMessage,
|
||||
}));
|
||||
@@ -82,17 +109,62 @@ function createStream(chunks: Array<{ result?: string; usage?: unknown }>) {
|
||||
};
|
||||
}
|
||||
|
||||
function flushAsyncTasks(): Promise<void> {
|
||||
return new Promise((resolve) => setTimeout(resolve, 0));
|
||||
function flushAsyncTasks(iterations = 1): Promise<void> {
|
||||
return new Promise((resolve) => {
|
||||
const next = (remaining: number) => {
|
||||
if (remaining <= 0) {
|
||||
resolve();
|
||||
return;
|
||||
}
|
||||
|
||||
setTimeout(() => next(remaining - 1), 0);
|
||||
};
|
||||
|
||||
next(iterations);
|
||||
});
|
||||
}
|
||||
|
||||
const spreadsheetCapability = {
|
||||
skillKey: 'minimax-xlsx',
|
||||
slug: 'minimax-xlsx',
|
||||
name: 'MiniMax XLSX',
|
||||
description: 'Analyze spreadsheet files such as .xlsx and .csv.',
|
||||
enabled: true,
|
||||
category: 'document',
|
||||
allowedTools: [],
|
||||
operationHints: ['read', 'analyze'],
|
||||
triggerHints: ['spreadsheet', 'excel'],
|
||||
inputExtensions: ['.xlsx', '.csv', '.tsv'],
|
||||
requiredEnvVars: [],
|
||||
requiresAuth: false,
|
||||
plannerSummary: 'document skill; operations: read, analyze; inputs: .xlsx, .csv, .tsv',
|
||||
renderHints: {
|
||||
card: 'document-analysis',
|
||||
preferredView: 'table',
|
||||
skillType: 'spreadsheet',
|
||||
},
|
||||
};
|
||||
|
||||
describe('chat runtime context', () => {
|
||||
beforeEach(() => {
|
||||
vi.resetModules();
|
||||
vi.clearAllMocks();
|
||||
mocks.sessionMessages.length = 0;
|
||||
mocks.activeRuns.clear();
|
||||
mocks.maybeHandleBrowserOpenMessage.mockReturnValue(false);
|
||||
mocks.maybeHandleSkillInstallMessage.mockReturnValue(false);
|
||||
mocks.getEnabledSkillCapabilities.mockReturnValue([]);
|
||||
mocks.providerChat.mockResolvedValue(createStream([{ result: 'done' }]));
|
||||
mocks.providerGetCapabilities.mockReturnValue({
|
||||
structuredMessages: false,
|
||||
toolCalls: false,
|
||||
toolResults: false,
|
||||
thinking: false,
|
||||
});
|
||||
mocks.toolRuntimeRun.mockReset();
|
||||
mocks.createChatToolRuntime.mockImplementation(() => ({
|
||||
run: mocks.toolRuntimeRun,
|
||||
}));
|
||||
});
|
||||
|
||||
it('prepends the zn-ai runtime context before provider chat runs', async () => {
|
||||
@@ -134,4 +206,212 @@ describe('chat runtime context', () => {
|
||||
content: expect.stringContaining('skills.install'),
|
||||
});
|
||||
});
|
||||
|
||||
it('persists tool_use -> tool_result -> final for planner-first spreadsheet execution', async () => {
|
||||
mocks.getEnabledSkillCapabilities.mockReturnValue([spreadsheetCapability]);
|
||||
mocks.providerChat.mockResolvedValue(createStream([{ result: 'Final answer from provider.' }]));
|
||||
mocks.toolRuntimeRun.mockImplementation(async (invocation: { toolCallId: string; toolName: string; input: unknown }) => {
|
||||
const payload = {
|
||||
ok: true,
|
||||
summary: 'Spreadsheet analysis completed.',
|
||||
structuredData: {
|
||||
reports: [{ filePath: 'C:\\tmp\\report.xlsx', rows: 3 }],
|
||||
},
|
||||
renderHints: {
|
||||
card: 'document-analysis',
|
||||
preferredView: 'table',
|
||||
skillType: 'spreadsheet',
|
||||
},
|
||||
raw: {
|
||||
reports: [{ filePath: 'C:\\tmp\\report.xlsx', rows: 3 }],
|
||||
},
|
||||
};
|
||||
|
||||
return {
|
||||
preflight: {
|
||||
ok: true,
|
||||
status: 'ready',
|
||||
toolCallId: invocation.toolCallId,
|
||||
toolName: invocation.toolName,
|
||||
normalizedInput: invocation.input,
|
||||
summary: 'Ready to analyze the spreadsheet.',
|
||||
},
|
||||
execution: {
|
||||
ok: true,
|
||||
status: 'completed',
|
||||
toolCallId: invocation.toolCallId,
|
||||
toolName: invocation.toolName,
|
||||
normalizedInput: invocation.input,
|
||||
summary: 'Spreadsheet analysis completed.',
|
||||
raw: payload.raw,
|
||||
durationMs: 12,
|
||||
},
|
||||
normalized: {
|
||||
ok: true,
|
||||
status: 'completed',
|
||||
toolCallId: invocation.toolCallId,
|
||||
toolName: invocation.toolName,
|
||||
summary: 'Spreadsheet analysis completed.',
|
||||
payload,
|
||||
block: {
|
||||
type: 'tool_result',
|
||||
toolCallId: invocation.toolCallId,
|
||||
content: 'Spreadsheet analysis completed.',
|
||||
result: payload,
|
||||
summary: 'Spreadsheet analysis completed.',
|
||||
ok: true,
|
||||
},
|
||||
transcriptMessage: {
|
||||
role: 'tool_result',
|
||||
content: [
|
||||
{
|
||||
type: 'tool_result',
|
||||
toolCallId: invocation.toolCallId,
|
||||
content: 'Spreadsheet analysis completed.',
|
||||
result: payload,
|
||||
summary: 'Spreadsheet analysis completed.',
|
||||
ok: true,
|
||||
},
|
||||
],
|
||||
timestamp: Date.now(),
|
||||
toolCallId: invocation.toolCallId,
|
||||
toolName: invocation.toolName,
|
||||
toolCall: {
|
||||
id: invocation.toolCallId,
|
||||
name: invocation.toolName,
|
||||
input: invocation.input,
|
||||
summary: 'Spreadsheet analysis completed.',
|
||||
},
|
||||
toolResult: payload,
|
||||
},
|
||||
},
|
||||
};
|
||||
});
|
||||
|
||||
const { handleChatSend } = await import('../electron/gateway/handlers/chat');
|
||||
const broadcast = vi.fn();
|
||||
|
||||
const result = handleChatSend(
|
||||
{
|
||||
sessionKey: 'agent:test:main',
|
||||
message: {
|
||||
role: 'user',
|
||||
content: 'Use minimax-xlsx to analyze this spreadsheet.',
|
||||
_attachedFiles: [
|
||||
{
|
||||
fileName: 'report.xlsx',
|
||||
mimeType: 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
|
||||
fileSize: 2048,
|
||||
preview: null,
|
||||
filePath: 'C:\\tmp\\report.xlsx',
|
||||
source: 'user-upload',
|
||||
},
|
||||
],
|
||||
},
|
||||
},
|
||||
broadcast,
|
||||
);
|
||||
|
||||
expect(result.runId).toBeTypeOf('string');
|
||||
expect(mocks.sessionMessages).toHaveLength(2);
|
||||
expect(mocks.sessionMessages[1]).toEqual(expect.objectContaining({
|
||||
role: 'assistant',
|
||||
toolName: 'minimax-xlsx',
|
||||
content: [
|
||||
expect.objectContaining({
|
||||
type: 'tool_use',
|
||||
name: 'minimax-xlsx',
|
||||
}),
|
||||
],
|
||||
}));
|
||||
|
||||
await flushAsyncTasks(4);
|
||||
|
||||
expect(mocks.toolRuntimeRun).toHaveBeenCalledWith(
|
||||
expect.objectContaining({
|
||||
toolName: 'minimax-xlsx',
|
||||
source: 'planner',
|
||||
}),
|
||||
expect.objectContaining({
|
||||
sessionKey: 'agent:test:main',
|
||||
runId: result.runId,
|
||||
files: expect.arrayContaining([
|
||||
expect.objectContaining({
|
||||
filePath: 'C:\\tmp\\report.xlsx',
|
||||
}),
|
||||
]),
|
||||
}),
|
||||
);
|
||||
expect(mocks.providerChat).toHaveBeenCalledTimes(1);
|
||||
|
||||
const [messages, model] = mocks.providerChat.mock.calls[0] ?? [];
|
||||
expect(model).toBe('gpt-4o-mini');
|
||||
expect(messages).toEqual(expect.arrayContaining([
|
||||
expect.objectContaining({
|
||||
role: 'system',
|
||||
content: expect.stringContaining('minimax-xlsx'),
|
||||
}),
|
||||
expect.objectContaining({
|
||||
role: 'assistant',
|
||||
content: [
|
||||
expect.objectContaining({
|
||||
type: 'tool_use',
|
||||
name: 'minimax-xlsx',
|
||||
}),
|
||||
],
|
||||
}),
|
||||
expect.objectContaining({
|
||||
role: 'tool_result',
|
||||
content: [
|
||||
expect.objectContaining({
|
||||
type: 'tool_result',
|
||||
summary: 'Spreadsheet analysis completed.',
|
||||
}),
|
||||
],
|
||||
}),
|
||||
]));
|
||||
|
||||
expect(mocks.sessionMessages.map((message) => message.role)).toEqual([
|
||||
'user',
|
||||
'assistant',
|
||||
'tool_result',
|
||||
'assistant',
|
||||
]);
|
||||
expect(mocks.sessionMessages[2]).toEqual(expect.objectContaining({
|
||||
role: 'tool_result',
|
||||
toolName: 'minimax-xlsx',
|
||||
toolResult: expect.objectContaining({
|
||||
summary: 'Spreadsheet analysis completed.',
|
||||
}),
|
||||
_toolStatuses: [
|
||||
expect.objectContaining({
|
||||
name: 'minimax-xlsx',
|
||||
status: 'completed',
|
||||
summary: 'Spreadsheet analysis completed.',
|
||||
}),
|
||||
],
|
||||
}));
|
||||
expect(mocks.sessionMessages[3]).toEqual(expect.objectContaining({
|
||||
role: 'assistant',
|
||||
content: 'Final answer from provider.',
|
||||
}));
|
||||
|
||||
expect(broadcast).toHaveBeenCalledWith(expect.objectContaining({
|
||||
type: 'tool:status',
|
||||
toolName: 'minimax-xlsx',
|
||||
status: 'running',
|
||||
}));
|
||||
expect(broadcast).toHaveBeenCalledWith(expect.objectContaining({
|
||||
type: 'tool:status',
|
||||
toolName: 'minimax-xlsx',
|
||||
status: 'completed',
|
||||
}));
|
||||
expect(broadcast).toHaveBeenLastCalledWith(expect.objectContaining({
|
||||
type: 'chat:final',
|
||||
runId: result.runId,
|
||||
message: expect.objectContaining({
|
||||
content: 'Final answer from provider.',
|
||||
}),
|
||||
}));
|
||||
});
|
||||
});
|
||||
|
||||
Reference in New Issue
Block a user