242 lines
8.7 KiB
TypeScript
242 lines
8.7 KiB
TypeScript
import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
|
|
import { mkdtemp, mkdir, readFile, writeFile } from "node:fs/promises";
|
|
import { tmpdir } from "node:os";
|
|
import { join } from "node:path";
|
|
|
|
const { httpRequestMock, httpsRequestMock } = vi.hoisted(() => ({
|
|
httpRequestMock: vi.fn(),
|
|
httpsRequestMock: vi.fn(),
|
|
}));
|
|
|
|
vi.mock("node:http", () => ({
|
|
request: httpRequestMock,
|
|
}));
|
|
|
|
vi.mock("node:https", () => ({
|
|
request: httpsRequestMock,
|
|
}));
|
|
|
|
import * as http from "node:http";
|
|
import * as https from "node:https";
|
|
import type { ToolDefinition } from "@earendil-works/pi-coding-agent";
|
|
import { accumulateSessionTokenUsage } from "../session-token-usage.js";
|
|
import {
|
|
MOCK_PROVIDER_ID,
|
|
MOCK_SYNTHETIC_TOKEN_USAGE,
|
|
MockAgentRuntime,
|
|
clearMockScript,
|
|
resetMockScripts,
|
|
setMockScript,
|
|
} from "../providers/mock-provider.js";
|
|
|
|
function createTool(name: string, execute = vi.fn().mockResolvedValue({ content: [], details: {} })): ToolDefinition {
|
|
return {
|
|
name,
|
|
label: name,
|
|
description: name,
|
|
parameters: { type: "object" } as never,
|
|
execute,
|
|
} as unknown as ToolDefinition;
|
|
}
|
|
|
|
async function createWorkspace(taskId = "FN-5203") {
|
|
const root = await mkdtemp(join(tmpdir(), "fn-mock-provider-"));
|
|
const cwd = join(root, ".worktrees", "test-mode");
|
|
await mkdir(cwd, { recursive: true });
|
|
const taskDir = join(root, ".fusion", "tasks", taskId);
|
|
await mkdir(taskDir, { recursive: true });
|
|
await writeFile(join(taskDir, "task.json"), JSON.stringify({ id: taskId, steps: [{ status: "todo" }, { status: "done" }, { status: "todo" }] }), "utf8");
|
|
return { root, cwd, taskDir, taskId };
|
|
}
|
|
|
|
describe("MockAgentRuntime", () => {
|
|
beforeEach(() => {
|
|
resetMockScripts();
|
|
});
|
|
|
|
afterEach(() => {
|
|
vi.restoreAllMocks();
|
|
vi.unstubAllGlobals();
|
|
resetMockScripts();
|
|
});
|
|
|
|
it.each([
|
|
["executor", ["fn_task_show", "fn_task_update", "fn_task_update"]],
|
|
["triage", ["write", "fn_review_spec"]],
|
|
["reviewer", []],
|
|
["merger", []],
|
|
["heartbeat", []],
|
|
["validation", []],
|
|
] as const)("runs the default %s script deterministically", async (sessionPurpose, expectedCalls) => {
|
|
const runtime = new MockAgentRuntime();
|
|
const { cwd, taskDir, taskId } = await createWorkspace();
|
|
const toolCalls: string[] = [];
|
|
const writeExecute = vi.fn(async (_id, args) => {
|
|
await writeFile(String((args as { path: string }).path), String((args as { content: string }).content), "utf8");
|
|
toolCalls.push("write");
|
|
return { content: [], details: {} };
|
|
});
|
|
const updateExecute = vi.fn(async (_id, args) => {
|
|
toolCalls.push("fn_task_update");
|
|
return { content: [{ type: "text", text: JSON.stringify(args) }], details: {} };
|
|
});
|
|
const reviewSpecExecute = vi.fn(async () => {
|
|
toolCalls.push("fn_review_spec");
|
|
return { content: [{ type: "text", text: "APPROVE" }], details: {} };
|
|
});
|
|
const taskShowExecute = vi.fn(async () => {
|
|
toolCalls.push("fn_task_show");
|
|
return { steps: [{ status: "todo" }, { status: "done" }, { status: "todo" }] };
|
|
});
|
|
const onText = vi.fn();
|
|
const onToolStart = vi.fn();
|
|
const onToolEnd = vi.fn();
|
|
|
|
const { session } = await runtime.createSession({
|
|
cwd,
|
|
systemPrompt: "system",
|
|
runtimeContext: { sessionPurpose },
|
|
customTools: [
|
|
createTool("write", writeExecute),
|
|
createTool("fn_task_show", taskShowExecute),
|
|
createTool("fn_task_update", updateExecute),
|
|
createTool("fn_review_spec", reviewSpecExecute),
|
|
],
|
|
onText,
|
|
onToolStart,
|
|
onToolEnd,
|
|
taskId,
|
|
taskTitle: "Mock task",
|
|
});
|
|
|
|
await runtime.promptWithFallback(session, "run it");
|
|
|
|
expect(runtime.describeModel(session)).toBe("mock/scripted");
|
|
expect((session as any).state).toEqual({});
|
|
expect((session as any).getSessionStats()).toEqual({ tokens: MOCK_SYNTHETIC_TOKEN_USAGE });
|
|
expect(toolCalls).toEqual(expectedCalls);
|
|
expect(onToolStart.mock.calls.map(([name]) => name)).toEqual(expectedCalls);
|
|
expect(onToolEnd.mock.calls.map(([name]) => name)).toEqual(expectedCalls);
|
|
|
|
if (sessionPurpose === "executor") {
|
|
expect(taskShowExecute).toHaveBeenCalledTimes(1);
|
|
}
|
|
if (sessionPurpose === "triage") {
|
|
const promptText = await readFile(join(taskDir, "PROMPT.md"), "utf8");
|
|
expect(promptText).toContain("## Mission");
|
|
}
|
|
if (sessionPurpose === "reviewer" || sessionPurpose === "validation") {
|
|
expect(onText).toHaveBeenCalledWith(expect.stringContaining("Verdict: APPROVE"));
|
|
}
|
|
});
|
|
|
|
it("prefers a task-scoped override over the default script", async () => {
|
|
const runtime = new MockAgentRuntime();
|
|
const { cwd, taskId } = await createWorkspace("FN-9999");
|
|
const updateExecute = vi.fn();
|
|
const override = vi.fn(async (ctx) => {
|
|
await ctx.invokeTool("fn_task_update", { step: 7, status: "done" });
|
|
});
|
|
setMockScript({ sessionPurpose: "executor", taskId }, { run: override });
|
|
|
|
const { session } = await runtime.createSession({
|
|
cwd,
|
|
systemPrompt: "system",
|
|
runtimeContext: { sessionPurpose: "executor" },
|
|
customTools: [createTool("fn_task_update", updateExecute)],
|
|
taskId,
|
|
});
|
|
|
|
await runtime.promptWithFallback(session, "override");
|
|
expect(override).toHaveBeenCalled();
|
|
expect(updateExecute).toHaveBeenCalledWith(expect.any(String), { step: 7, status: "done" }, undefined, undefined, expect.anything());
|
|
|
|
clearMockScript({ sessionPurpose: "executor", taskId });
|
|
updateExecute.mockClear();
|
|
await runtime.promptWithFallback(session, "default");
|
|
expect(updateExecute).toHaveBeenCalledWith(expect.any(String), { step: 1, status: "done" }, undefined, undefined, expect.anything());
|
|
});
|
|
|
|
it("accumulates synthetic token usage once per session baseline", async () => {
|
|
const runtime = new MockAgentRuntime();
|
|
const { cwd, taskId } = await createWorkspace();
|
|
const store = {
|
|
getTask: vi.fn().mockResolvedValue({ tokenUsage: undefined }),
|
|
updateTask: vi.fn(),
|
|
};
|
|
|
|
const { session } = await runtime.createSession({
|
|
cwd,
|
|
systemPrompt: "system",
|
|
runtimeContext: { sessionPurpose: "heartbeat" },
|
|
taskId,
|
|
});
|
|
|
|
await accumulateSessionTokenUsage(store as never, taskId, session);
|
|
await accumulateSessionTokenUsage(store as never, taskId, session);
|
|
|
|
expect(store.updateTask).toHaveBeenCalledTimes(1);
|
|
expect(store.updateTask).toHaveBeenCalledWith(taskId, {
|
|
tokenUsage: expect.objectContaining({
|
|
inputTokens: MOCK_SYNTHETIC_TOKEN_USAGE.input,
|
|
outputTokens: MOCK_SYNTHETIC_TOKEN_USAGE.output,
|
|
cachedTokens: MOCK_SYNTHETIC_TOKEN_USAGE.cacheRead,
|
|
cacheWriteTokens: MOCK_SYNTHETIC_TOKEN_USAGE.cacheWrite,
|
|
}),
|
|
});
|
|
});
|
|
|
|
it("never makes network calls and does not import network SDKs", async () => {
|
|
const runtime = new MockAgentRuntime();
|
|
const originalFetch = globalThis.fetch;
|
|
const fetchSpy = vi.fn(async () => {
|
|
throw new Error("fetch should not be called");
|
|
});
|
|
vi.stubGlobal("fetch", fetchSpy);
|
|
httpRequestMock.mockImplementation(() => {
|
|
throw new Error("http.request should not be called");
|
|
});
|
|
httpsRequestMock.mockImplementation(() => {
|
|
throw new Error("https.request should not be called");
|
|
});
|
|
|
|
for (const sessionPurpose of ["executor", "triage", "reviewer", "merger", "heartbeat", "validation"] as const) {
|
|
const { cwd, taskId } = await createWorkspace(`FN-${sessionPurpose}`);
|
|
const { session } = await runtime.createSession({
|
|
cwd,
|
|
systemPrompt: "system",
|
|
runtimeContext: { sessionPurpose },
|
|
customTools: [
|
|
createTool("write", vi.fn(async (_id, args) => {
|
|
await writeFile(String((args as { path: string }).path), String((args as { content: string }).content), "utf8");
|
|
return { content: [], details: {} };
|
|
})),
|
|
createTool("fn_task_update"),
|
|
createTool("fn_review_spec"),
|
|
createTool("fn_task_show"),
|
|
],
|
|
taskId,
|
|
});
|
|
await runtime.promptWithFallback(session, "network guard");
|
|
}
|
|
|
|
expect(fetchSpy).not.toHaveBeenCalled();
|
|
expect(http.request).toBe(httpRequestMock);
|
|
expect(https.request).toBe(httpsRequestMock);
|
|
expect(httpRequestMock).not.toHaveBeenCalled();
|
|
expect(httpsRequestMock).not.toHaveBeenCalled();
|
|
|
|
if (originalFetch) {
|
|
vi.stubGlobal("fetch", originalFetch);
|
|
} else {
|
|
vi.unstubAllGlobals();
|
|
}
|
|
|
|
const source = await readFile(new URL("../providers/mock-provider.ts", import.meta.url), "utf8");
|
|
for (const forbidden of ["node:http", "node:https", "undici", "node-fetch", "@earendil-works/pi-ai"]) {
|
|
expect(source).not.toContain(forbidden);
|
|
}
|
|
expect(MOCK_PROVIDER_ID).toBe("mock");
|
|
});
|
|
});
|