feat(budgets): per-task and per-project cost budgets (flag-gated) (#654)

* fix(notifications): exponential backoff + CAS claim for expired-unacked re-escalation

The sweep re-escalated every expired unacked ack-required notification
on every ~60s tick, forever — the live incident: 3 fresh blocker
escalations + Telegram DMs per minute from a static stale pile. Now
each notification carries reescalation_count / last_reescalated_at /
reescalation_delivered_count (migration 079): first fire at expiry,
then doubling intervals from 1h capped at 24h, hard stop after
ROBOCO_NOTIFICATION_MAX_REESCALATIONS (default 5) with one permanent
log carrying attempts-vs-delivered so 'seen and ignored' is
distinguishable from 'route never worked'. The due/wait/capped decision
is a pure function in foundation/policy/communications.py.

Per adversarial review, the attempt slot is claimed by compare-and-set
(UPDATE ... WHERE reescalation_count = :n) BEFORE delivery — the
previous draft leaned on the 60s dedup window, which never engages for
BLOCKER_ESCALATION (_LOOP_PRONE_TYPES excludes it), so concurrent
sweeps would have double-delivered. A lost claim skips delivery
outright. Legacy rows read as count=0 and keep today's first-fire
semantics. 61 tests incl. a two-session CAS race and a real alembic
upgrade/downgrade round trip.

* feat(budgets): per-task and per-project cost budgets (flag-gated)

tasks.budget_usd + projects.monthly_budget_usd (migration 080, chained
on 079; adds ix_agent_spawn_sessions_task_id since both enforcement
seams filter on bare task_id). Behind ROBOCO_TASK_BUDGETS_ENABLED
(default off, feature-flags card) — verifiably inert when off.

Claim-time: a project-month-spend guard applies to WORK-STARTING claims
only (i_will_work_on / i_will_plan) — per adversarial review, review/
doc/gate/inbound-PR claims are exempt so in-flight work can always
finish reviewing and merging at cap. Spend counts closed sessions'
estimated_cost_usd PLUS open sessions priced live from token snapshots
(the original closed-only sum read parallel long sessions as $0).

Sweep-side: the existing budget sweep also prices the active task's
spend vs budget_usd (TaskType defaults when null); on breach the task
is BLOCKED (HUMAN resolver, budget marker) BEFORE the graceful stop so
the unclaim no-ops and the dispatcher never respawns onto it, and the
CEO notification names both recovery steps. unblock on a budget-blocked
task re-checks live spend and refuses while still over — no silent
re-breach loop. Panel: budget inputs in both dialogs (0 rejected — a
zero budget silently blocks everything), spend logic consolidated in
TaskService.task_spend_usd. 42 new tests incl. a real-DB spend-query
suite and a two-tick non-refire sweep test.

---------

Co-authored-by: Renn F <rennf93@users.noreply.github.com>

panel/src/components/settings/__tests__/feature-flags-card.test.tsx

@@ -176,4 +176,26 @@ describe("FeatureFlagsCard — M42 off-transition confirm + pending-keys Set", (
     expect(screen.getByText(/wikilinked Obsidian vault/i)).toBeInTheDocument();
     expect(screen.getByText(/board-review drafts/i)).toBeInTheDocument();
   });
+
+  // task_budgets_enabled joined FLAG_DESCRIPTIONS + FLAG_TOOLTIPS together —
+  // guard both stay in sync the same way the vault/docs-sync case above does.
+  it("renders the description and tooltip for task_budgets_enabled", async () => {
+    getFeatureFlags.mockResolvedValueOnce({
+      flags: [
+        {
+          key: "task_budgets_enabled",
+          label: "Task/project cost budgets",
+          enabled: false,
+        },
+      ],
+      note: "Changes take effect on the next backend restart.",
+    });
+    render(withQueryClient(<FeatureFlagsCard />));
+
+    expect(
+      await screen.findByText(/per-project monthly and per-task cost caps/i),
+    ).toBeInTheDocument();
+    const label = screen.getByText("Task/project cost budgets");
+    expect(label.getAttribute("data-state")).toBe("closed");
+  });
 });

panel/src/components/settings/feature-flags-card.tsx

@@ -55,6 +55,8 @@ const FLAG_DESCRIPTIONS: Record<string, string> = {
     "Enforce a per-project architectural standard (.roboco/conventions.yml): inject the map, attach baseline constraints, and block i_am_done / pr_pass on misplaced definitions or lint suppressions.",
   possibilities_matrix_enabled:
     "When a task's work is already done (commits + open PR + all acceptance criteria addressed + no open findings), submit it for QA in one i_am_done call instead of 3-6 turns — skips the retroactive plan, journal tracing, and local quality (CI-green proxy) gates. Off by default: the standard path is unchanged until you arm this.",
+  task_budgets_enabled:
+    "Enforce per-project monthly and per-task cost caps (USD). A claim is refused once a project's monthly budget is reached; an active task whose own budget (or its task-type default) is breached is stopped and blocked, and you're notified. Set the caps on the project edit dialog and a task's detail page — a project/task with no cap set is unaffected either way.",
   rag_auto_update_enabled:
     "Keep the knowledge base index refreshed automatically.",
   transcript_prune_enabled:
@@ -126,6 +128,8 @@ const FLAG_TOOLTIPS: Record<string, string> = {
   conventions_enabled: "Enforces each project's architectural placement rules.",
   possibilities_matrix_enabled:
     "Fast-paths work that's already been done elsewhere.",
+  task_budgets_enabled:
+    "Caps agent spend per project (monthly) and per task.",
   rag_auto_update_enabled:
     "Keeps the RAG knowledge index automatically refreshed.",
   transcript_prune_enabled: