feat: session goals - server-driven goal loop with independent small-model audit (#2148)

Arm the target button in the composer and the next prompt becomes a goal:
the server keeps the session working toward it (idle tick -> small-model
audit -> continuation) until the objective is verifiably complete, blocked,
or out of budget — even with the UI closed.

Server (packages/web/server/lib/session-goal):
- event-driven loop on the global SSE hub; goal state lives in
  session.metadata.openchamber.goal (merge-safe patches, stale-write guard
  by goal id), so it survives restarts and syncs to every client for free
- the small-model audit (objective + last assistant turn only, language
  pinned to the objective) is the sole termination authority; blocked needs
  3 consecutive verdicts, audit outages tolerate one unaudited continuation
  then stop the goal as resumable-blocked
- hard stops: optional token budget, auto-continuation cap (Resume grants a
  fresh allowance), turn errors; user abort pauses the goal instead of
  blocking it, and resuming over an aborted tail nudges immediately
- token accounting as a snapshot of the latest turn (input + cache.read +
  output), goal-relative via a creation baseline and segmented across
  compactions; a compaction summary skips the audit and continues
- continuations reuse the session's own provider/model/agent/variant

UI:
- three-mode target button (arm / disarm / manage dialog), informational
  goal strip with inline pause/resume and an Evaluating indicator, sidebar
  state glyph, objective length counter (2000-char server clamp),
  read-only completed goals
- goal entry points: composer (sessions and drafts), start-new-session-
  from-answer dialog, plan implement dialog (plan content becomes the
  objective), scheduled tasks (Run as goal + budget)
- Settings -> Chat -> Goal: feature toggle + default token budget with
  three-layer parity (web server, client persistence, VS Code bridge);
  VS Code renders goal state but hides the entry points (the loop runs in
  the web server only)

Notifications: per-turn "ready" notifications are suppressed while a goal
is active; settling sends one final notification (desktop, web-push, APNs
generic titles with the session name as body) honoring the completion
toggle. Error/question/permission notifications are untouched.

Docs: user guide (session-goals) in all 9 locales + sidebar entry,
scheduled-tasks cross-reference, server module DOCUMENTATION.md.
This commit is contained in:
Bohdan Triapitsyn
2026-07-12 01:23:22 +03:00
committed by GitHub
parent 82c039117a
commit bb45164ae8
73 changed files with 3330 additions and 29 deletions
@@ -6,6 +6,7 @@ import { Input } from '@/components/ui/input';
import { Textarea } from '@/components/ui/textarea';
import { Select, SelectContent, SelectItem, SelectTrigger, SelectValue } from '@/components/ui/select';
import { Checkbox } from '@/components/ui/checkbox';
import { NumberInput } from '@/components/ui/number-input';
import { Button } from '@/components/ui/button';
import { MobileOverlayPanel } from '@/components/ui/MobileOverlayPanel';
import { toast } from '@/components/ui';
@@ -466,6 +467,8 @@ type ScheduledTaskDraft = {
modelID: string;
variant: string;
agent: string;
goalEnabled: boolean;
goalTokenBudget: number | null;
};
state?: ScheduledTask['state'];
};
@@ -515,6 +518,8 @@ const toDraft = (
modelID: defaults.modelID,
variant: defaults.variant,
agent: defaults.agent,
goalEnabled: false,
goalTokenBudget: null,
},
};
}
@@ -548,6 +553,10 @@ const toDraft = (
modelID: task.execution.modelID,
variant: task.execution.variant || '',
agent: task.execution.agent || '',
goalEnabled: task.execution.goalEnabled === true,
goalTokenBudget: typeof task.execution.goalTokenBudget === 'number' && task.execution.goalTokenBudget > 0
? task.execution.goalTokenBudget
: null,
},
state: task.state,
};
@@ -1153,6 +1162,10 @@ export function ScheduledTaskEditorDialog(props: {
modelID: draft.execution.modelID,
...(draft.execution.variant.trim() ? { variant: draft.execution.variant.trim() } : {}),
...(draft.execution.agent.trim() ? { agent: draft.execution.agent.trim() } : {}),
...(draft.execution.goalEnabled ? { goalEnabled: true } : {}),
...(draft.execution.goalEnabled && draft.execution.goalTokenBudget
? { goalTokenBudget: draft.execution.goalTokenBudget }
: {}),
},
...(draft.state ? { state: draft.state } : {}),
};
@@ -1608,6 +1621,49 @@ export function ScheduledTaskEditorDialog(props: {
) : null}
</div>
</div>
<div className="flex flex-wrap items-center gap-x-8 gap-y-2">
<label className="inline-flex cursor-pointer items-center gap-2">
<Checkbox
checked={draft.execution.goalEnabled}
onChange={(goalEnabled) => setDraft((prev) => ({
...prev,
execution: { ...prev.execution, goalEnabled },
}))}
ariaLabel={t('sessions.scheduledTasks.editor.goal.aria')}
/>
<span className="typography-meta">{t('sessions.scheduledTasks.editor.goal.label')}</span>
</label>
{draft.execution.goalEnabled ? (
<label className="inline-flex cursor-pointer items-center gap-2">
<Checkbox
checked={draft.execution.goalTokenBudget !== null}
onChange={(hasBudget) => setDraft((prev) => ({
...prev,
execution: { ...prev.execution, goalTokenBudget: hasBudget ? 200_000 : null },
}))}
ariaLabel={t('sessions.scheduledTasks.editor.goal.budgetAria')}
/>
<span className="typography-meta">{t('sessions.scheduledTasks.editor.goal.budgetLabel')}</span>
</label>
) : null}
{draft.execution.goalEnabled && draft.execution.goalTokenBudget !== null ? (
<NumberInput
value={draft.execution.goalTokenBudget}
onValueChange={(value) => {
if (typeof value === 'number' && Number.isFinite(value) && value > 0) {
setDraft((prev) => ({
...prev,
execution: { ...prev.execution, goalTokenBudget: Math.floor(value) },
}));
}
}}
min={1000}
max={100000000}
step={50000}
/>
) : null}
</div>
</div>
);