agenthub/tests/architectBudget.test.ts
chahinebrini 6d088648da feat(hub): Korrektheits- und Realtime-Härtung — Alias, Lifecycle, Presence, Architekten-Pulse (TSK-0242/0245/0249)
Drei vom Architekten abgenommene Tasks, gebündelt als Checkpoint:

- TSK-0242: Agent-Alias-Mapping (kimi-ah → kimi kanonisiert, Rollen → preferredAgent),
  reopenTask räumt claimedBy ab, fsWatch reindiziert direkte Datei-Edits,
  work-Default 300s → 50s, task_list mit Limit.
- TSK-0245: zwei Agent-Klassen (dispatch loop|architect). Watchdog mahnt
  architekt-getriebene Agenten nur noch EINMAL statt im Minutentakt;
  `task dispatch` startet sie explizit, `task record` trägt extern
  erledigte Arbeit mit origin=external nach.
- TSK-0249: Lifecycle wird serverseitig erzwungen (open→review scheitert mit
  klarer Meldung), claimedBy/doneBy überleben bis done, Presence pro Agent,
  Review-Watchdog, GET /architect/pulse (1.4 kB statt 34 kB, since-Cursor,
  omitted statt stillem Abschneiden), unbekannter Agent → 400 statt 500.

Alle Punkte live am laufenden Hub nachgemessen, nicht aus Agenten-Logs übernommen.
Tests: 242 → 279 grün, tsc sauber.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-07-29 17:48:24 +02:00

60 lines
2.8 KiB
TypeScript

import { describe, it, expect, beforeEach, afterEach } from 'vitest';
import { mkdtempSync, rmSync } from 'fs';
import { tmpdir } from 'os';
import { join } from 'path';
import { init } from '../src/cli/commands/init.js';
import { createTask, claimTask, reviewTask, doneTask } from '../src/core/services/taskService.js';
import { createHandoff } from '../src/core/services/handoffService.js';
import { createMessage } from '../src/core/services/messageService.js';
import { computeBudget, TOKENS_PER_COORDINATION_ACTION } from '../src/core/services/budgetService.js';
/** Architect coordination work must be visible in the budget report (TSK-0226). */
describe('budget: architect activity', () => {
let cwd: string;
beforeEach(() => {
cwd = mkdtempSync(join(tmpdir(), 'ah-budget-'));
init(cwd, { projectName: 'budget-test', yes: true });
});
afterEach(() => {
rmSync(cwd, { recursive: true, force: true });
});
it('counts reviews, handoffs and messages as architect activity', () => {
// Implementer-owned task; the architect only coordinates around it.
createTask(cwd, { title: 'Feature', role: 'implementer', assignedTo: 'codex' });
claimTask(cwd, 'TSK-0001', 'codex');
reviewTask(cwd, 'TSK-0001', 'claude');
createHandoff(cwd, { fromRole: 'architect', toRole: 'implementer', fromAgent: 'claude', toAgent: 'codex', taskId: 'TSK-0001', summary: 'scope' });
createMessage(cwd, { from: 'claude', to: 'codex', text: 'please review' });
const report = computeBudget(cwd);
const claude = report.agents.find((a) => a.name === 'claude');
expect(claude).toBeDefined();
// reviewer + handoff + message = 3 actions, all estimated (never silent fabrication).
expect(claude!.actions).toBe(3);
expect(claude!.tokens).toBe(3 * TOKENS_PER_COORDINATION_ACTION);
expect(claude!.estimated).toBe(true);
expect(claude!.estimatedTokens).toBe(claude!.tokens);
expect(claude!.costEur).toBeGreaterThan(0);
});
it('a done approval by someone other than the owner counts as an action', () => {
createTask(cwd, { title: 'Feature', role: 'implementer', assignedTo: 'codex' });
claimTask(cwd, 'TSK-0001', 'codex');
reviewTask(cwd, 'TSK-0001', 'claude');
// Architect approves: doneBy = claude, owner = codex.
doneTask(cwd, 'TSK-0001', { doneBy: 'claude', doneTokens: 8000 });
const report = computeBudget(cwd);
const claude = report.agents.find((a) => a.name === 'claude')!;
// review + approval; the implementer keeps the real task tokens.
expect(claude.actions).toBe(2);
const codex = report.agents.find((a) => a.name === 'codex')!;
expect(codex.realTokens).toBe(8000);
// No double count: claude's action estimate does not include the 8000.
expect(claude.tokens).toBe(2 * TOKENS_PER_COORDINATION_ACTION);
});
});