refactor(schema): make runs.task_id nullable, delete heartbeat workaround

The NOT NULL constraint forced heartbeat runs to synthesize a fake task
row (`[Heartbeat] X — Check for work`) so the FK could resolve. Three
layers of code — acquireHeartbeatTask, findReusableHeartbeatTask,
cleanupDuplicateHeartbeatTasks, plus the is_heartbeat flag column + UI
filter — existed to paper over one bad constraint. Remove the
constraint and the whole workaround vanishes.

SQLite table rebuild with data preservation. Live production DB, no
backup — migration is atomic, idempotent, and verified before DROP.

## Migration v35 (rebuild runs with nullable task_id)

Uses `fn:` form so it can:
- Snapshot `COUNT(*) FROM runs` before copy
- Walk `PRAGMA table_info(runs)` to detect drift columns (v34 declared
  a column set; any additive ALTER since v14 shows up here)
- Assert `COUNT(*) FROM runs_new == beforeCount` before `DROP TABLE runs`
- `PRAGMA defer_foreign_keys = ON` so run_events(run_id) FK defers to
  commit time, by which runs has the same id set under the new name
- Recreate all 6 indexes on the new table

If the assertion fails, the migration throws inside the existing
`transaction()` wrapper. SQLite rolls back the whole migration, old
`runs` table is untouched, `schema_migrations` never records v35. Next
app start retries cleanly.

## Migration v36 (drop tasks.is_heartbeat + synthetic rows)

Runs now carry `task_id = NULL` for heartbeat dispatches. The
`is_heartbeat` flag is dead data going forward.

1. `UPDATE runs SET task_id = NULL WHERE task_id IN (SELECT id FROM tasks WHERE is_heartbeat = 1)`
2. `DELETE FROM tasks WHERE is_heartbeat = 1` (synthetic rows, not real work)
3. Rebuild `tasks` without the `is_heartbeat` column (same defer-FK +
   row-count-assert pattern as v35)

## Code cleanup

- `RunRecord.taskId: string | null`; `createRun` / `cancelRunForDeletion`
  / `finishRun` guard every `UPDATE tasks` path with a null-check.
- `TaskRecord.isHeartbeat` removed; every SELECT / filter / UI reference
  dropped.
- Delete `acquireHeartbeatTask` / `findReusableHeartbeatTask` /
  `cleanupDuplicateHeartbeatTasks` — dead after the constraint lifts.
- Delete `src/main/orchestration/heartbeat-task.ts` (pure workaround
  constant file).
- `heartbeat.ts` creates runs with `taskId: null` when the agent has no
  assigned work. Turn context already supports nullable task
  (heartbeatFallbackSection); prompt surface unchanged.
- `validateTaskCreation` drops the `isHeartbeat` parameter — user-
  authored + automation-authored tasks are never heartbeat.
- Renderer: `run.taskId ? taskMap.get(...) : undefined` in Dashboard /
  RunCenter / RunDetail / RunQueue.
- Delete `tests/unit/heartbeat-coalescing.test.ts` (12 tests that
  exercised the deleted DB methods).
- Drop 1 isHeartbeat-bypass test from `task-quality-gates.test.ts`.

## Grep verification

- `isHeartbeatTaskTitle src/` → 0
- `acquireHeartbeatTask src/` → migration comment only
- `is_heartbeat src/` → schema.ts migration history only
- `isHeartbeat src/` → 0

Tests: 578/578 (was 591 — delta is 13 tests that tested deleted code).
This commit is contained in:
ZacharyZhang-NY committed 2026-04-17 16:16:08 -04:00
1 parent 92795f7bfb
commit 7e1d4199ee
44 files changed
+385 -674

No files matched your search

+3 -150
View File
@@ -1,4 +1,3 @@
import { randomUUID } from "node:crypto";
import type { DatabaseSync } from "node:sqlite";
import type { TaskRecord } from "@shared/types";
import {
@@ -12,7 +11,7 @@ import { chooseBestTaskAssignee } from "./task-assignment";
export function buildTaskStatusMethods(
getDb: () => DatabaseSync,
readAll: ReadAll,
transaction: Transaction,
_transaction: Transaction,
deps: {
addComment: (input: { companyId: string; taskId: string; authorAgentId: string | null; authorName: string; body: string }) => void;
addActivity: (input: { companyId: string; actor: string; action: string; entityType: string; entityId: string; detail: string }) => void;
@@ -102,152 +101,6 @@ export function buildTaskStatusMethods(
return { previousStatus };
},
findReusableHeartbeatTask(agentId: string, companyId: string): TaskRecord | null {
const row = getDb()
.prepare(
`SELECT id, company_id as companyId, project_id as projectId, goal_id as goalId,
parent_id as parentId, title, description,
assignee_agent_id as assigneeAgentId, workspace_id as workspaceId,
status, priority, active_run_id as activeRunId,
session_json as sessionJson, session_display_id as sessionDisplayId,
cost_usd as costUsd, requires_user_review as requiresUserReview,
is_heartbeat as isHeartbeat,
created_at as createdAt, updated_at as updatedAt
FROM tasks
WHERE company_id = ? AND assignee_agent_id = ? AND is_heartbeat = 1
AND status IN ('done', 'in_review')
ORDER BY updated_at DESC, rowid DESC
LIMIT 1`,
)
.get(companyId, agentId) as TaskRecord | undefined;
if (row) {
row.requiresUserReview = Boolean(row.requiresUserReview);
row.isHeartbeat = Boolean(row.isHeartbeat);
}
return row ?? null;
},
acquireHeartbeatTask(input: {
agentId: string;
companyId: string;
runId: string;
workspaceId: string | null;
title: string;
description: string;
priority: string;
}): TaskRecord | null {
return transaction<TaskRecord | null>(() => {
const ts = nowIso();
const db = getDb();
const claimable = db
.prepare(
`SELECT id, status FROM tasks
WHERE company_id = ? AND assignee_agent_id = ? AND is_heartbeat = 1
AND status IN ('todo', 'backlog', 'done', 'in_review')
AND checkout_run_id IS NULL
ORDER BY
CASE status WHEN 'todo' THEN 0 WHEN 'backlog' THEN 1 WHEN 'in_review' THEN 2 WHEN 'done' THEN 3 ELSE 4 END ASC,
updated_at DESC, rowid DESC
LIMIT 1`,
)
.get(input.companyId, input.agentId) as { id: string; status: string } | undefined;
let taskId: string | null = null;
if (claimable) {
const claimResult = db
.prepare(
`UPDATE tasks
SET checkout_run_id = ?, execution_locked_at = ?, status = 'in_progress',
started_at = COALESCE(started_at, ?), completed_at = NULL, cancelled_at = NULL,
updated_at = ?
WHERE id = ? AND checkout_run_id IS NULL
AND status IN ('todo', 'backlog', 'done', 'in_review')`,
)
.run(input.runId, ts, ts, ts, claimable.id);
if (claimResult.changes > 0) {
taskId = claimable.id;
} else {
return null;
}
} else {
const inflight = db
.prepare(
`SELECT 1 FROM tasks
WHERE company_id = ? AND assignee_agent_id = ? AND is_heartbeat = 1
AND status NOT IN ('done', 'cancelled')
LIMIT 1`,
)
.get(input.companyId, input.agentId);
if (inflight) return null;
taskId = randomUUID();
db.prepare(
`INSERT INTO tasks
(id, company_id, project_id, goal_id, parent_id, title, description,
assignee_agent_id, workspace_id, status, priority,
checkout_run_id, execution_locked_at, started_at, is_heartbeat,
created_at, updated_at)
VALUES (?, ?, NULL, NULL, NULL, ?, ?, ?, ?, 'in_progress', ?, ?, ?, ?, 1, ?, ?)`,
).run(
taskId,
input.companyId,
input.title,
input.description,
input.agentId,
input.workspaceId,
input.priority,
input.runId,
ts,
ts,
ts,
ts,
);
}
const row = db
.prepare(
`SELECT id, company_id as companyId, project_id as projectId, goal_id as goalId,
parent_id as parentId, title, description,
assignee_agent_id as assigneeAgentId, workspace_id as workspaceId,
status, priority, active_run_id as activeRunId,
session_json as sessionJson, session_display_id as sessionDisplayId,
cost_usd as costUsd, requires_user_review as requiresUserReview,
is_heartbeat as isHeartbeat, task_type as taskType,
started_at as startedAt, completed_at as completedAt, cancelled_at as cancelledAt,
created_at as createdAt, updated_at as updatedAt
FROM tasks WHERE id = ?`,
)
.get(taskId) as TaskRecord | undefined;
if (!row) return null;
row.requiresUserReview = Boolean(row.requiresUserReview);
row.isHeartbeat = Boolean(row.isHeartbeat);
row.startedAt = row.startedAt ?? null;
row.completedAt = row.completedAt ?? null;
row.cancelledAt = row.cancelledAt ?? null;
return row;
});
},
cleanupDuplicateHeartbeatTasks(companyId: string): number {
const ts = nowIso();
const result = getDb()
.prepare(
`UPDATE tasks SET status = 'cancelled', updated_at = ?
WHERE company_id = ? AND is_heartbeat = 1
AND status IN ('done', 'in_review', 'todo', 'blocked', 'backlog')
AND id NOT IN (
SELECT id FROM (
SELECT id, ROW_NUMBER() OVER (PARTITION BY assignee_agent_id ORDER BY updated_at DESC, rowid DESC) as rn
FROM tasks
WHERE company_id = ? AND is_heartbeat = 1 AND status != 'cancelled'
) WHERE rn = 1
)`,
)
.run(ts, companyId, companyId);
return Number(result.changes);
},
listReviewTasksForManager(managerId: string, companyId: string): TaskRecord[] {
return readAll<TaskRecord>(
`select t.id, t.company_id as companyId, t.project_id as projectId, t.goal_id as goalId,
@@ -255,11 +108,11 @@ export function buildTaskStatusMethods(
t.assignee_agent_id as assigneeAgentId, t.workspace_id as workspaceId,
t.status, t.priority, t.active_run_id as activeRunId,
t.session_json as sessionJson, t.session_display_id as sessionDisplayId,
t.cost_usd as costUsd, t.is_heartbeat as isHeartbeat,
t.cost_usd as costUsd,
t.created_at as createdAt, t.updated_at as updatedAt
from tasks t
inner join agents a on t.assignee_agent_id = a.id
where t.company_id = ? and t.status = 'in_review' and a.reports_to = ? and t.is_heartbeat = 0
where t.company_id = ? and t.status = 'in_review' and a.reports_to = ?
order by ${taskPriorityOrderSql("t.priority")} desc, t.created_at asc`,
companyId,
managerId,