agents-relay 1.0.48 → 1.0.49
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/dist/cli.js
CHANGED
|
@@ -328,7 +328,7 @@ async function appendPlannedTask(store, job, task) {
|
|
|
328
328
|
const decision = orchestratorPlanningDecision(job, [task], decisionId, task.parentTaskId, job.tasks.filter(item => item.kind === 'planner').length, 'orchestrator');
|
|
329
329
|
if (!job.tasks.some(item => item.id === decision.id))
|
|
330
330
|
await store.appendTask(decision);
|
|
331
|
-
const planned = { ...task, parentTaskId:
|
|
331
|
+
const planned = { ...task, parentTaskId: decision.id, planningDecisionId: decision.id, planningSource: 'orchestrator', planningRound: decision.planningRound };
|
|
332
332
|
await store.appendTask(planned);
|
|
333
333
|
job.tasks.push(planned, decision);
|
|
334
334
|
return planned;
|
package/package.json
CHANGED
|
@@ -12,3 +12,15 @@ A hotfix is a fast path for a live blocking problem:
|
|
|
12
12
|
4. The orchestrator completes the normal post-merge/publish reconciliation and cleanup.
|
|
13
13
|
|
|
14
14
|
Do not create a blocking job/PR before fixing the live runtime. Keep the hotfix minimal and do not leave the live patch as the only durable implementation.
|
|
15
|
+
|
|
16
|
+
## P0 / self job workflow
|
|
17
|
+
|
|
18
|
+
A **P0 job** (alias: **self job**) is executed directly by the current active orchestrator thread:
|
|
19
|
+
|
|
20
|
+
1. The orchestrator creates the durable job.
|
|
21
|
+
2. The orchestrator immediately creates the required orchestrator-owned task(s); do not leave the P0 job idle.
|
|
22
|
+
3. The current orchestrator thread starts executing those tasks immediately rather than delegating them to another worker.
|
|
23
|
+
4. After implementation, the orchestrator performs/reconciles review against the current head.
|
|
24
|
+
5. If review passes and merge requirements are satisfied, the orchestrator merges the job itself and completes normal post-merge reconciliation.
|
|
25
|
+
|
|
26
|
+
`job create` still only creates durable job state; the P0 workflow requires the orchestrator to follow it immediately with task creation and execution.
|
|
@@ -40,18 +40,6 @@ A normal task never merges a PR. After all non-superseded work gates pass, Relay
|
|
|
40
40
|
|
|
41
41
|
Final-review routing follows the same worker-selection boundary as executable work. Capability, availability, explicit reviewer constraints, and resource health are hard gates. Among eligible workers, Relay prefers a reviewer with a different worker identity from the implementation tasks; when that is unavailable it prefers a different provider/model, then falls back to the first eligible worker. The selected worker, `selectedBy` reason, and `reviewerRouting` event data are durable evidence. No provider is a universal reviewer default.
|
|
42
42
|
|
|
43
|
-
## P0 / self job workflow
|
|
44
|
-
|
|
45
|
-
A **P0 job** (alias: **self job**) is executed directly by the current active orchestrator thread:
|
|
46
|
-
|
|
47
|
-
1. The orchestrator creates the durable job.
|
|
48
|
-
2. The orchestrator immediately creates the required orchestrator-owned task(s); do not leave the P0 job idle.
|
|
49
|
-
3. The current orchestrator thread starts executing those tasks immediately rather than delegating them to another worker.
|
|
50
|
-
4. After implementation, the orchestrator performs/reconciles review against the current head.
|
|
51
|
-
5. If review passes and merge requirements are satisfied, the orchestrator merges the job itself and completes normal post-merge reconciliation.
|
|
52
|
-
|
|
53
|
-
`job create` still only creates durable job state; the P0 workflow requires the orchestrator to follow it immediately with task creation and execution.
|
|
54
|
-
|
|
55
43
|
## Fixed and autonomous jobs
|
|
56
44
|
|
|
57
45
|
`job create` persists only the durable job. It never creates executable work. The top-level orchestrator owns the initial execution decision and, when the user intends the job to start, submits the first task batch separately. A zero-task autonomous job stays idle and reconciliation must not bootstrap its first planner.
|