Evenfire is an intelligence layer platform for running LLM agents that take real actions, on your infrastructure
TypeScript
import { beforeEach, describe, expect, it, vi } from 'vitest'
import type { Pool } from 'pg'
import { type SchedulingRecipe, reconcileScheduling } from './schedulingHandler'
function makeRecipe(overrides: Partial<SchedulingRecipe> = {}): SchedulingRecipe {
return {
metadata: {
name: 'daily-report',
namespace: 'sandbox-recipes',
uid: 'uid-123',
labels: { 'clerum.io/workflow-team-id': '11111111-1111-4111-8111-111111111111' },
...(overrides.metadata ?? {}),
},
spec: {
scheduling: {
cron: '0 9 * * *',
timezone: 'UTC',
},
...(overrides.spec ?? {}),
},
}
}
describe('reconcileScheduling', () => {
const query = vi.fn()
const pool = { query } as unknown as Pool
beforeEach(() => {
vi.clearAllMocks()
})
it('denormalizes runRetention on schedule creation', async () => {
query
.mockResolvedValueOnce({ rows: [], rowCount: 0 })
.mockResolvedValueOnce({ rows: [], rowCount: 1 })
const result = await reconcileScheduling(
pool,
makeRecipe({
spec: {
scheduling: {
cron: '0 9 * * *',
timezone: 'UTC',
},
runRetention: {
maxRunDurationSeconds: 7200,
ttlSecondsAfterFinished: 3600,
},
},
})
)
expect(result).toEqual({ action: 'created' })
const insertCall = query.mock.calls[1] as [string, unknown[]]
expect(insertCall[0]).toContain('INSERT INTO workflow_schedules')
expect(insertCall[0]).toContain('max_duration_seconds')
expect(insertCall[0]).toContain('ttl_seconds_after_finished')
expect(insertCall[1]?.[8]).toBe(7200)
expect(insertCall[1]?.[9]).toBe(3600)
})
it('defaults schedule artifact retention to 30 days when omitted', async () => {
query
.mockResolvedValueOnce({ rows: [], rowCount: 0 })
.mockResolvedValueOnce({ rows: [], rowCount: 1 })
const result = await reconcileScheduling(pool, makeRecipe())
expect(result).toEqual({ action: 'created' })
const insertCall = query.mock.calls[1] as [string, unknown[]]
expect(insertCall[0]).toContain('ttl_seconds_after_finished')
expect(insertCall[1]?.[9]).toBe(2_592_000)
})
it('updates the denormalized retention fields when retention changes', async () => {
query
.mockResolvedValueOnce({
rows: [
{
cron_expression: '0 9 * * *',
timezone: 'UTC',
enabled: true,
allowed_actors: null,
max_duration_seconds: 600,
ttl_seconds_after_finished: 1200,
},
],
rowCount: 1,
})
.mockResolvedValueOnce({ rows: [], rowCount: 1 })
const result = await reconcileScheduling(
pool,
makeRecipe({
spec: {
scheduling: {
cron: '0 9 * * *',
timezone: 'UTC',
},
runRetention: {
maxRunDurationSeconds: 900,
ttlSecondsAfterFinished: 1800,
},
},
})
)
expect(result).toEqual({ action: 'updated' })
const updateCall = query.mock.calls[1] as [string, unknown[]]
expect(updateCall[0]).toContain('UPDATE workflow_schedules')
expect(updateCall[0]).toContain('max_duration_seconds = $4')
expect(updateCall[0]).toContain('ttl_seconds_after_finished = $5')
expect(updateCall[0]).not.toContain('next_fire_at =')
expect(updateCall[1]?.[3]).toBe(900)
expect(updateCall[1]?.[4]).toBe(1800)
})
})