import { randomUUID } from "node:crypto"; import { tmpdir } from "node:os"; import { join } from "node:path"; import { describe, expect, test } from "vitest"; import { DatabaseConnection } from "../../src/main/database/connection"; import { OnboardingService } from "../../src/main/services/onboarding-service"; import { ProviderService } from "../../src/main/services/provider-service"; import { providerAdapters } from "../../src/main/services/provider-adapters"; import { TaskExecutionService } from "../../src/main/services/task-execution-service"; import { claimNextQueuedRun } from "../../src/main/app/task-queue-worker"; import { requestsForRun } from "../../src/main/services/provider-request-store"; import type { AgentTurnResult, ProviderToolTurnInput } from "../../src/main/services/provider-tool-turn"; import { answeringModelCheck } from "../helpers/model-check"; const testVault = { encryptSecret: (secret: string) => Buffer.from(secret, "utf8"), decryptSecret: (ciphertext: Buffer) => ciphertext.toString("utf8"), }; describe("a run's model calls", () => { test("every call a run made is listed in order with its purpose, including the turn that failed", async () => { let turns = 0; const turn = async (_input: ProviderToolTurnInput): Promise => { turns += 1; if (turns === 1) return { text: "Checking in.", toolCalls: [{ id: "c1", name: "company_status", arguments: {} }], rawResponseJson: "{}" }; throw new Error("Provider overloaded"); }; const connection = new DatabaseConnection(join(tmpdir(), `ac-run-requests-${randomUUID()}.db`)); const providers = new ProviderService(connection.db, providerAdapters, testVault, answeringModelCheck(turn)); const onboarding = new OnboardingService({ db: connection.db, saveProviderCredential: (input) => providers.saveCredential(input) }); const execution = new TaskExecutionService(connection.db, providers); let orgId = ""; connection.transaction(() => { orgId = onboarding.bootstrap({ companyName: "FlowPilot", companyDescription: "CSM analytics.", employeeName: "Yang", workspacePath: "", executionMode: "ai_ceo", autoApproveHires: true, reviewDeliverables: false, goalTitle: "Ship the MVP.", goalDescription: "A working product.", providerKey: "openai", providerSecret: "sk-test-key", providerModel: "gpt-5.4", }).orgId; }); const runId = claimNextQueuedRun(connection)!.id; await expect(execution.executeRun(orgId, runId)).rejects.toThrow(); // Each call carries its own output, so the run's detail does not depend on a snapshot window. const turnCall = requestsForRun(connection.db, orgId, runId).find((request) => request.purpose === "agent_turn" && request.state === "succeeded")!; expect([typeof turnCall.responseText, typeof turnCall.inputCharCount, typeof turnCall.outputCharCount]).toEqual(["string", "number", "number"]); expect(requestsForRun(connection.db, orgId, runId).map((request) => [request.purpose, request.state])).toEqual([ ["model_check", "succeeded"], ["agent_turn", "succeeded"], ["agent_turn", "failed"], ]); connection.close(); }); });