@hecer/yoke 1.1.0 → 1.2.0

package/bench/fixtures/routing-queue/package.json

@@ -0,0 +1,9 @@
+{
+  "name": "bench-routing-queue",
+  "private": true,
+  "version": "0.0.0",
+  "type": "module",
+  "scripts": {
+    "test": "node bench-verify.mjs"
+  }
+}

package/bench/fixtures/routing-queue/src/task-queue.mjs

@@ -0,0 +1 @@
+// Benchmark fixture: implementation intentionally left for the coding agent.

package/bench/fixtures/routing-queue/tests/STORY-1.test.mjs

@@ -0,0 +1,68 @@
+import test from 'node:test'
+import assert from 'node:assert/strict'
+import { TaskQueue } from '../src/task-queue.mjs'
+
+const makeQueue = () => {
+  let now = 1_000
+  const queue = new TaskQueue({ clock: () => now, maxAttempts: 2, baseDelayMs: 100 })
+  return { queue, setNow: value => { now = value } }
+}
+
+test('enqueue validates type and creates stable immutable snapshots', () => {
+  const { queue } = makeQueue()
+  assert.throws(() => queue.enqueue({ type: '', payload: {} }), /type/i)
+  const task = queue.enqueue({ type: 'email', payload: { to: 'a@example.test' } })
+  assert.equal(task.status, 'queued')
+  assert.equal(task.priority, 0)
+  assert.equal(task.createdAt, 1_000)
+  assert.throws(() => { task.status = 'dead' }, TypeError)
+  task.payload.to = 'changed@example.test'
+  assert.equal(queue.get(task.id).payload.to, 'a@example.test')
+})
+
+test('idempotency keys return the original task without growing the queue', () => {
+  const { queue } = makeQueue()
+  const first = queue.enqueue({ type: 'email', payload: { n: 1 }, idempotencyKey: 'order-7' })
+  const second = queue.enqueue({ type: 'email', payload: { n: 2 }, idempotencyKey: 'order-7' })
+  assert.equal(second.id, first.id)
+  assert.equal(second.payload.n, 1)
+  assert.equal(queue.list().length, 1)
+})
+
+test('claim selects priority first and FIFO for ties', () => {
+  const { queue, setNow } = makeQueue()
+  const low = queue.enqueue({ type: 'low', payload: {}, priority: 1 })
+  setNow(1_001)
+  const highFirst = queue.enqueue({ type: 'high-a', payload: {}, priority: 5 })
+  setNow(1_002)
+  const highSecond = queue.enqueue({ type: 'high-b', payload: {}, priority: 5 })
+  assert.equal(queue.claim('worker-1', { leaseMs: 50 }).id, highFirst.id)
+  assert.equal(queue.claim('worker-2', { leaseMs: 50 }).id, highSecond.id)
+  assert.equal(queue.claim('worker-3', { leaseMs: 50 }).id, low.id)
+  assert.equal(queue.claim('worker-4'), null)
+})
+
+test('claim records lease ownership and complete enforces it', () => {
+  const { queue } = makeQueue()
+  const original = queue.enqueue({ type: 'work', payload: {} })
+  const claimed = queue.claim('worker-a', { leaseMs: 250 })
+  assert.equal(claimed.attempts, 1)
+  assert.equal(claimed.workerId, 'worker-a')
+  assert.equal(claimed.leaseUntil, 1_250)
+  assert.throws(() => queue.complete(original.id, 'worker-b', 42), /worker|owner|lease/i)
+  const done = queue.complete(original.id, 'worker-a', { ok: true })
+  assert.equal(done.status, 'completed')
+  assert.deepEqual(done.result, { ok: true })
+  assert.throws(() => queue.complete(original.id, 'worker-a', {}), /state|active|completed/i)
+})
+
+test('get/list snapshots cannot mutate internal state', () => {
+  const { queue } = makeQueue()
+  const task = queue.enqueue({ type: 'work', payload: { nested: { value: 1 } } })
+  const listed = queue.list()
+  listed[0].payload.nested.value = 9
+  listed.length = 0
+  assert.equal(queue.get(task.id).payload.nested.value, 1)
+  assert.equal(queue.list().length, 1)
+  assert.equal(queue.get('missing'), null)
+})

package/bench/fixtures/routing-queue/tests/STORY-2.test.mjs

@@ -0,0 +1,72 @@
+import test from 'node:test'
+import assert from 'node:assert/strict'
+import { TaskQueue } from '../src/task-queue.mjs'
+
+const makeQueue = (options = {}) => {
+  let now = 1_000
+  const queue = new TaskQueue({ clock: () => now, maxAttempts: 2, baseDelayMs: 100, ...options })
+  return { queue, now: () => now, advance: ms => { now += ms } }
+}
+
+test('fail requeues with exponential backoff and respects availability', () => {
+  const { queue, now, advance } = makeQueue({ maxAttempts: 3 })
+  const task = queue.enqueue({ type: 'work', payload: {} })
+  queue.claim('worker')
+  const first = queue.fail(task.id, 'worker', new Error('temporary'))
+  assert.equal(first.status, 'queued')
+  assert.equal(first.availableAt, now() + 100)
+  assert.equal(queue.claim('early'), null)
+  advance(100)
+  queue.claim('worker')
+  const second = queue.fail(task.id, 'worker', 'again')
+  assert.equal(second.availableAt, now() + 200)
+})
+
+test('failure at maxAttempts moves a task to dead letters', () => {
+  const { queue, advance } = makeQueue({ maxAttempts: 2 })
+  const task = queue.enqueue({ type: 'work', payload: {} })
+  queue.claim('worker')
+  queue.fail(task.id, 'worker', 'first')
+  advance(100)
+  queue.claim('worker')
+  const dead = queue.fail(task.id, 'worker', new Error('final'))
+  assert.equal(dead.status, 'dead')
+  assert.match(dead.error, /final/)
+  assert.equal(queue.claim('other'), null)
+})
+
+test('reapExpired requeues only expired active leases', () => {
+  const { queue, advance } = makeQueue()
+  const first = queue.enqueue({ type: 'a', payload: {}, priority: 2 })
+  queue.enqueue({ type: 'b', payload: {}, priority: 1 })
+  queue.claim('worker-a', { leaseMs: 50 })
+  queue.claim('worker-b', { leaseMs: 500 })
+  advance(51)
+  assert.equal(queue.reapExpired(), 1)
+  assert.equal(queue.get(first.id).status, 'queued')
+  assert.equal(queue.claim('worker-c').id, first.id)
+})
+
+test('list filters by status and type and stats reports every state', () => {
+  const { queue } = makeQueue({ maxAttempts: 1 })
+  const completed = queue.enqueue({ type: 'email', payload: {} })
+  queue.claim('a')
+  queue.complete(completed.id, 'a', 'ok')
+  const dead = queue.enqueue({ type: 'email', payload: {} })
+  queue.claim('b')
+  queue.fail(dead.id, 'b', 'bad')
+  queue.enqueue({ type: 'report', payload: {} })
+  assert.equal(queue.list({ type: 'email' }).length, 2)
+  assert.equal(queue.list({ status: 'dead' }).length, 1)
+  assert.deepEqual(queue.stats(), { total: 3, queued: 1, active: 0, completed: 1, dead: 1 })
+})
+
+test('invalid identifiers, owners, and transitions are rejected', () => {
+  const { queue } = makeQueue()
+  assert.throws(() => queue.claim(''), /worker/i)
+  assert.throws(() => queue.fail('missing', 'worker', 'x'), /task|missing|unknown/i)
+  const task = queue.enqueue({ type: 'work', payload: {} })
+  assert.throws(() => queue.fail(task.id, 'worker', 'x'), /state|active|queued/i)
+  queue.claim('owner')
+  assert.throws(() => queue.fail(task.id, 'intruder', 'x'), /worker|owner|lease/i)
+})

package/bench/results/routing-queue-codex-routing-off-2026-08-01T21-44-00.json

@@ -0,0 +1,46 @@
+{
+  "schemaVersion": 1,
+  "fixtureVersion": "routing-queue@1",
+  "runner": "codex",
+  "sampleLabel": "routing-ab-controlled-baseline",
+  "permissionProfile": "unsafe",
+  "yokeVersion": "1.1.1",
+  "fixture": "routing-queue",
+  "routing": "off",
+  "startedAt": "2026-08-01T21:44:00.963Z",
+  "wallClockMs": 299787,
+  "exitCode": 0,
+  "finalState": "complete",
+  "verdict": "completed",
+  "blocker": null,
+  "conflicts": 0,
+  "iterations": 2,
+  "finalTestsPass": true,
+  "progress": {
+    "passed": 2,
+    "total": 2
+  },
+  "usageAvailable": true,
+  "modelAvailable": true,
+  "requestedParentModel": "gpt-5.6-sol",
+  "tokens": {
+    "inputTokens": 579301,
+    "outputTokens": 9103
+  },
+  "modelCalls": [],
+  "stories": [
+    {
+      "id": "STORY-1",
+      "durationMs": 173400,
+      "iterations": 1,
+      "finalTestsPass": true
+    },
+    {
+      "id": "STORY-2",
+      "durationMs": 125693,
+      "iterations": 1,
+      "finalTestsPass": true
+    }
+  ],
+  "srcLoc": 133
+}

package/bench/results/routing-queue-codex-routing-on-2026-08-01T21-50-31.json

@@ -0,0 +1,109 @@
+{
+  "schemaVersion": 1,
+  "fixtureVersion": "routing-queue@1",
+  "runner": "codex",
+  "sampleLabel": "routing-ab-controlled-routed",
+  "permissionProfile": "unsafe",
+  "yokeVersion": "1.1.1",
+  "fixture": "routing-queue",
+  "routing": "on",
+  "startedAt": "2026-08-01T21:50:32.408Z",
+  "wallClockMs": 283134,
+  "exitCode": 0,
+  "finalState": "complete",
+  "verdict": "completed",
+  "blocker": null,
+  "conflicts": 0,
+  "iterations": 2,
+  "finalTestsPass": true,
+  "progress": {
+    "passed": 2,
+    "total": 2
+  },
+  "usageAvailable": true,
+  "modelAvailable": true,
+  "requestedParentModel": "gpt-5.6-sol",
+  "tokens": {
+    "inputTokens": 542788,
+    "outputTokens": 5486,
+    "calls": [
+      {
+        "role": "orchestrator",
+        "provider": "codex",
+        "requestedModel": "gpt-5.6-sol",
+        "inputTokens": 19065,
+        "outputTokens": 28,
+        "durationMs": 22124
+      },
+      {
+        "role": "worker",
+        "provider": "codex",
+        "inputTokens": 272894,
+        "outputTokens": 3201,
+        "durationMs": 124464
+      },
+      {
+        "role": "orchestrator",
+        "provider": "codex",
+        "requestedModel": "gpt-5.6-sol",
+        "inputTokens": 19046,
+        "outputTokens": 30,
+        "durationMs": 19208
+      },
+      {
+        "role": "worker",
+        "provider": "codex",
+        "inputTokens": 231783,
+        "outputTokens": 2227,
+        "durationMs": 102089
+      }
+    ]
+  },
+  "modelCalls": [
+    {
+      "role": "orchestrator",
+      "provider": "codex",
+      "requestedModel": "gpt-5.6-sol",
+      "inputTokens": 19065,
+      "outputTokens": 28,
+      "durationMs": 22124
+    },
+    {
+      "role": "worker",
+      "provider": "codex",
+      "inputTokens": 272894,
+      "outputTokens": 3201,
+      "durationMs": 124464
+    },
+    {
+      "role": "orchestrator",
+      "provider": "codex",
+      "requestedModel": "gpt-5.6-sol",
+      "inputTokens": 19046,
+      "outputTokens": 30,
+      "durationMs": 19208
+    },
+    {
+      "role": "worker",
+      "provider": "codex",
+      "inputTokens": 231783,
+      "outputTokens": 2227,
+      "durationMs": 102089
+    }
+  ],
+  "stories": [
+    {
+      "id": "STORY-1",
+      "durationMs": 148000,
+      "iterations": 1,
+      "finalTestsPass": true
+    },
+    {
+      "id": "STORY-2",
+      "durationMs": 122288,
+      "iterations": 1,
+      "finalTestsPass": true
+    }
+  ],
+  "srcLoc": 112
+}

package/bench/results/yoke-codex-study-codex-routing-off-2026-08-02T07-58-56.json

@@ -0,0 +1,65 @@
+{
+  "schemaVersion": 1,
+  "fixtureVersion": "yoke-codex-study@1",
+  "runner": "codex",
+  "sampleLabel": "codex-only-pair1-off",
+  "permissionProfile": "unsafe",
+  "yokeVersion": "1.1.1",
+  "fixture": "yoke-codex-study",
+  "routing": "off",
+  "startedAt": "2026-08-02T07:58:58.592Z",
+  "wallClockMs": 561860,
+  "exitCode": 0,
+  "finalState": "complete",
+  "verdict": "completed",
+  "blocker": null,
+  "conflicts": 0,
+  "iterations": 2,
+  "finalTestsPass": true,
+  "progress": {
+    "passed": 2,
+    "total": 2
+  },
+  "usageAvailable": true,
+  "modelAvailable": true,
+  "requestedParentModel": "gpt-5.6-sol",
+  "tokens": {
+    "inputTokens": 3255382,
+    "cachedInputTokens": 3050240,
+    "cacheWriteInputTokens": 0,
+    "outputTokens": 17468,
+    "reasoningOutputTokens": 7383
+  },
+  "modelCalls": [],
+  "tokenBreakdown": {
+    "inputTokens": 3255382,
+    "cachedInputTokens": 3050240,
+    "freshInputTokens": 205142,
+    "cacheWriteInputTokens": 0,
+    "outputTokens": 17468,
+    "reasoningOutputTokens": 7383
+  },
+  "stories": [
+    {
+      "id": "STUDY-1",
+      "durationMs": 345519,
+      "iterations": 1,
+      "finalTestsPass": true
+    },
+    {
+      "id": "STUDY-2",
+      "durationMs": 215666,
+      "iterations": 1,
+      "finalTestsPass": true
+    }
+  ],
+  "srcLoc": 5274,
+  "sourceFiles": 61,
+  "finalVerification": {
+    "buildExitCode": 0,
+    "testExitCode": 0,
+    "finalTestCount": 580,
+    "originalAcceptanceTestsExitCode": 0,
+    "originalAcceptanceTestCount": 6
+  }
+}

package/bench/results/yoke-codex-study-codex-routing-off-2026-08-02T08-08-48.json

@@ -0,0 +1,65 @@
+{
+  "schemaVersion": 1,
+  "fixtureVersion": "yoke-codex-study@1",
+  "runner": "codex",
+  "sampleLabel": "codex-only-pair2-off",
+  "permissionProfile": "unsafe",
+  "yokeVersion": "1.1.1",
+  "fixture": "yoke-codex-study",
+  "routing": "off",
+  "startedAt": "2026-08-02T08:08:51.009Z",
+  "wallClockMs": 579139,
+  "exitCode": 0,
+  "finalState": "complete",
+  "verdict": "completed",
+  "blocker": null,
+  "conflicts": 0,
+  "iterations": 2,
+  "finalTestsPass": true,
+  "progress": {
+    "passed": 2,
+    "total": 2
+  },
+  "usageAvailable": true,
+  "modelAvailable": true,
+  "requestedParentModel": "gpt-5.6-sol",
+  "tokens": {
+    "inputTokens": 2027874,
+    "cachedInputTokens": 1887744,
+    "cacheWriteInputTokens": 0,
+    "outputTokens": 14662,
+    "reasoningOutputTokens": 6087
+  },
+  "modelCalls": [],
+  "tokenBreakdown": {
+    "inputTokens": 2027874,
+    "cachedInputTokens": 1887744,
+    "freshInputTokens": 140130,
+    "cacheWriteInputTokens": 0,
+    "outputTokens": 14662,
+    "reasoningOutputTokens": 6087
+  },
+  "stories": [
+    {
+      "id": "STUDY-1",
+      "durationMs": 346078,
+      "iterations": 1,
+      "finalTestsPass": true
+    },
+    {
+      "id": "STUDY-2",
+      "durationMs": 232165,
+      "iterations": 1,
+      "finalTestsPass": true
+    }
+  ],
+  "srcLoc": 5259,
+  "sourceFiles": 61,
+  "finalVerification": {
+    "buildExitCode": 0,
+    "testExitCode": 0,
+    "finalTestCount": 578,
+    "originalAcceptanceTestsExitCode": 0,
+    "originalAcceptanceTestCount": 6
+  }
+}

package/bench/results/yoke-codex-study-codex-routing-off-2026-08-02T08-34-45.json

@@ -0,0 +1,65 @@
+{
+  "schemaVersion": 1,
+  "fixtureVersion": "yoke-codex-study@1",
+  "runner": "codex",
+  "sampleLabel": "codex-only-pair3-off",
+  "permissionProfile": "unsafe",
+  "yokeVersion": "1.1.1",
+  "fixture": "yoke-codex-study",
+  "routing": "off",
+  "startedAt": "2026-08-02T08:34:47.242Z",
+  "wallClockMs": 589501,
+  "exitCode": 0,
+  "finalState": "complete",
+  "verdict": "completed",
+  "blocker": null,
+  "conflicts": 0,
+  "iterations": 2,
+  "finalTestsPass": true,
+  "progress": {
+    "passed": 2,
+    "total": 2
+  },
+  "usageAvailable": true,
+  "modelAvailable": true,
+  "requestedParentModel": "gpt-5.6-sol",
+  "tokens": {
+    "inputTokens": 2279164,
+    "cachedInputTokens": 2147584,
+    "cacheWriteInputTokens": 0,
+    "outputTokens": 16008,
+    "reasoningOutputTokens": 6536
+  },
+  "modelCalls": [],
+  "tokenBreakdown": {
+    "inputTokens": 2279164,
+    "cachedInputTokens": 2147584,
+    "freshInputTokens": 131580,
+    "cacheWriteInputTokens": 0,
+    "outputTokens": 16008,
+    "reasoningOutputTokens": 6536
+  },
+  "stories": [
+    {
+      "id": "STUDY-1",
+      "durationMs": 376161,
+      "iterations": 1,
+      "finalTestsPass": true
+    },
+    {
+      "id": "STUDY-2",
+      "durationMs": 212311,
+      "iterations": 1,
+      "finalTestsPass": true
+    }
+  ],
+  "srcLoc": 5253,
+  "sourceFiles": 61,
+  "finalVerification": {
+    "buildExitCode": 0,
+    "testExitCode": 0,
+    "finalTestCount": 580,
+    "originalAcceptanceTestsExitCode": 0,
+    "originalAcceptanceTestCount": 6
+  }
+}

package/bench/results/yoke-codex-study-codex-routing-on-2026-08-02T07-51-59.json

@@ -0,0 +1,168 @@
+{
+  "schemaVersion": 1,
+  "fixtureVersion": "yoke-codex-study@1",
+  "runner": "codex",
+  "sampleLabel": "codex-only-pair1-on",
+  "permissionProfile": "unsafe",
+  "yokeVersion": "1.1.1",
+  "fixture": "yoke-codex-study",
+  "routing": "on",
+  "startedAt": "2026-08-02T07:52:01.670Z",
+  "wallClockMs": 383099,
+  "exitCode": 0,
+  "finalState": "complete",
+  "verdict": "completed",
+  "blocker": null,
+  "conflicts": 0,
+  "iterations": 2,
+  "finalTestsPass": true,
+  "progress": {
+    "passed": 2,
+    "total": 2
+  },
+  "usageAvailable": true,
+  "modelAvailable": true,
+  "requestedParentModel": "gpt-5.6-sol",
+  "tokens": {
+    "inputTokens": 1561371,
+    "cachedInputTokens": 1436672,
+    "cacheWriteInputTokens": 0,
+    "outputTokens": 9340,
+    "reasoningOutputTokens": 2010,
+    "calls": [
+      {
+        "role": "orchestrator",
+        "provider": "codex",
+        "requestedModel": "gpt-5.6-sol",
+        "requestedReasoningEffort": "low",
+        "inputTokens": 19164,
+        "cachedInputTokens": 11008,
+        "cacheWriteInputTokens": 0,
+        "outputTokens": 30,
+        "reasoningOutputTokens": 0,
+        "durationMs": 12390
+      },
+      {
+        "role": "worker",
+        "provider": "codex",
+        "profile": "codex-luna",
+        "requestedModel": "gpt-5.6-luna",
+        "requestedReasoningEffort": "low",
+        "inputTokens": 570526,
+        "cachedInputTokens": 528640,
+        "cacheWriteInputTokens": 0,
+        "outputTokens": 4335,
+        "reasoningOutputTokens": 343,
+        "durationMs": 171211
+      },
+      {
+        "role": "orchestrator",
+        "provider": "codex",
+        "requestedModel": "gpt-5.6-sol",
+        "requestedReasoningEffort": "low",
+        "inputTokens": 19161,
+        "cachedInputTokens": 11008,
+        "cacheWriteInputTokens": 0,
+        "outputTokens": 32,
+        "reasoningOutputTokens": 0,
+        "durationMs": 12716
+      },
+      {
+        "role": "worker",
+        "provider": "codex",
+        "profile": "codex-luna",
+        "requestedModel": "gpt-5.6-luna",
+        "requestedReasoningEffort": "low",
+        "inputTokens": 952520,
+        "cachedInputTokens": 886016,
+        "cacheWriteInputTokens": 0,
+        "outputTokens": 4943,
+        "reasoningOutputTokens": 1667,
+        "durationMs": 182866
+      }
+    ]
+  },
+  "modelCalls": [
+    {
+      "role": "orchestrator",
+      "provider": "codex",
+      "requestedModel": "gpt-5.6-sol",
+      "requestedReasoningEffort": "low",
+      "inputTokens": 19164,
+      "cachedInputTokens": 11008,
+      "cacheWriteInputTokens": 0,
+      "outputTokens": 30,
+      "reasoningOutputTokens": 0,
+      "durationMs": 12390
+    },
+    {
+      "role": "worker",
+      "provider": "codex",
+      "profile": "codex-luna",
+      "requestedModel": "gpt-5.6-luna",
+      "requestedReasoningEffort": "low",
+      "inputTokens": 570526,
+      "cachedInputTokens": 528640,
+      "cacheWriteInputTokens": 0,
+      "outputTokens": 4335,
+      "reasoningOutputTokens": 343,
+      "durationMs": 171211
+    },
+    {
+      "role": "orchestrator",
+      "provider": "codex",
+      "requestedModel": "gpt-5.6-sol",
+      "requestedReasoningEffort": "low",
+      "inputTokens": 19161,
+      "cachedInputTokens": 11008,
+      "cacheWriteInputTokens": 0,
+      "outputTokens": 32,
+      "reasoningOutputTokens": 0,
+      "durationMs": 12716
+    },
+    {
+      "role": "worker",
+      "provider": "codex",
+      "profile": "codex-luna",
+      "requestedModel": "gpt-5.6-luna",
+      "requestedReasoningEffort": "low",
+      "inputTokens": 952520,
+      "cachedInputTokens": 886016,
+      "cacheWriteInputTokens": 0,
+      "outputTokens": 4943,
+      "reasoningOutputTokens": 1667,
+      "durationMs": 182866
+    }
+  ],
+  "tokenBreakdown": {
+    "inputTokens": 1561371,
+    "cachedInputTokens": 1436672,
+    "freshInputTokens": 124699,
+    "cacheWriteInputTokens": 0,
+    "outputTokens": 9340,
+    "reasoningOutputTokens": 2010
+  },
+  "stories": [
+    {
+      "id": "STUDY-1",
+      "durationMs": 184822,
+      "iterations": 1,
+      "finalTestsPass": true
+    },
+    {
+      "id": "STUDY-2",
+      "durationMs": 197354,
+      "iterations": 1,
+      "finalTestsPass": true
+    }
+  ],
+  "srcLoc": 5252,
+  "sourceFiles": 61,
+  "finalVerification": {
+    "buildExitCode": 0,
+    "testExitCode": 0,
+    "finalTestCount": 578,
+    "originalAcceptanceTestsExitCode": 0,
+    "originalAcceptanceTestCount": 6
+  }
+}