feat(apify): digest rate cap (1 per artist per 24h) + email_send_log audit trail

lib/apify/digest/__tests__/maybeSendScrapeDigest.test.ts

@@ -3,6 +3,8 @@ import { maybeSendScrapeDigest } from "@/lib/apify/digest/maybeSendScrapeDigest"
 import { selectApifyScraperRunsByBatch } from "@/lib/supabase/apify_scraper_runs/selectApifyScraperRunsByBatch";
 import { getScrapeDigestRecipients } from "@/lib/apify/digest/getScrapeDigestRecipients";
 import { sendScrapeDigestEmail } from "@/lib/apify/digest/sendScrapeDigestEmail";
+import { selectRecentScrapeDigestLogs } from "@/lib/supabase/email_send_log/selectRecentScrapeDigestLogs";
+import { logEmailAttempt } from "@/lib/emails/logEmailAttempt";
 
 vi.mock("@/lib/supabase/apify_scraper_runs/selectApifyScraperRunsByBatch", () => ({
   selectApifyScraperRunsByBatch: vi.fn(),
@@ -13,6 +15,10 @@ vi.mock("@/lib/apify/digest/getScrapeDigestRecipients", () => ({
 vi.mock("@/lib/apify/digest/sendScrapeDigestEmail", () => ({
   sendScrapeDigestEmail: vi.fn(),
 }));
+vi.mock("@/lib/supabase/email_send_log/selectRecentScrapeDigestLogs", () => ({
+  selectRecentScrapeDigestLogs: vi.fn(async () => []),
+}));
+vi.mock("@/lib/emails/logEmailAttempt", () => ({ logEmailAttempt: vi.fn(async () => {}) }));
 
 const run = (over: Record<string, unknown>) => ({
   run_id: "r1",
@@ -27,7 +33,10 @@ const run = (over: Record<string, unknown>) => ({
 
 beforeEach(() => {
   vi.clearAllMocks();
-  vi.mocked(getScrapeDigestRecipients).mockResolvedValue(["owner@example.com"]);
+  vi.mocked(getScrapeDigestRecipients).mockResolvedValue({
+    emails: ["owner@example.com"],
+    artistIds: ["artist-1"],
+  } as never);
   vi.mocked(sendScrapeDigestEmail).mockResolvedValue({ id: "email-1" } as never);
 });
 
@@ -55,6 +64,21 @@ describe("maybeSendScrapeDigest", () => {
       { platform: "tiktok", postUrls: ["https://tiktok.com/v/2"] },
     ]); // x omitted — nothing new
     expect(arg.emails).toEqual(["owner@example.com"]);
+    // audit trail: the send is recorded per watched artist entity
+    expect(logEmailAttempt).toHaveBeenCalledWith(
+      expect.objectContaining({ accountId: "artist-1", status: "sent" }),
+    );
+  });
+
+  it("suppresses the digest when one was sent for the artist within 24h (rate cap)", async () => {
+    vi.mocked(selectApifyScraperRunsByBatch).mockResolvedValue([
+      run({ new_post_urls: ["https://instagram.com/p/1"] }),
+    ] as never);
+    vi.mocked(selectRecentScrapeDigestLogs).mockResolvedValue([
+      { account_id: "artist-1", created_at: "2026-07-07T00:00:00Z" },
+    ] as never);
+    expect(await maybeSendScrapeDigest("b1")).toBeNull();
+    expect(sendScrapeDigestEmail).not.toHaveBeenCalled();
   });
 
   it("sends nothing when the batch completes with zero new posts anywhere", async () => {

lib/apify/digest/getScrapeDigestRecipients.ts

@@ -7,9 +7,11 @@ import selectAccountEmails from "@/lib/supabase/account_emails/selectAccountEmai
  * of every artist watching any of them (same chain the per-platform alert
  * used). Recipients span tenants — senders must BCC (chat#1855).
  */
-export async function getScrapeDigestRecipients(socialIds: string[]): Promise<string[]> {
+export async function getScrapeDigestRecipients(
+  socialIds: string[],
+): Promise<{ emails: string[]; artistIds: string[] }> {
   const uniqueSocialIds = Array.from(new Set(socialIds.filter(Boolean)));
-  if (!uniqueSocialIds.length) return [];
+  if (!uniqueSocialIds.length) return { emails: [], artistIds: [] };
 
   const accountSocials = (
     await Promise.all(
@@ -24,5 +26,8 @@ export async function getScrapeDigestRecipients(socialIds: string[]): Promise<st
     new Set(accountArtistIds.map(a => a.account_id).filter((id): id is string => Boolean(id))),
   );
   const accountEmails = await selectAccountEmails({ accountIds: uniqueAccountIds });
-  return Array.from(new Set(accountEmails.map(e => e.email).filter(Boolean)));
+  return {
+    emails: Array.from(new Set(accountEmails.map(e => e.email).filter(Boolean))),
+    artistIds: Array.from(new Set(accountSocials.map(a => a.account_id).filter(Boolean))),
+  };
 }

lib/apify/digest/maybeSendScrapeDigest.ts

@@ -1,6 +1,8 @@
 import { selectApifyScraperRunsByBatch } from "@/lib/supabase/apify_scraper_runs/selectApifyScraperRunsByBatch";
 import { getScrapeDigestRecipients } from "@/lib/apify/digest/getScrapeDigestRecipients";
 import { sendScrapeDigestEmail } from "@/lib/apify/digest/sendScrapeDigestEmail";
+import { selectRecentScrapeDigestLogs } from "@/lib/supabase/email_send_log/selectRecentScrapeDigestLogs";
+import { logEmailAttempt } from "@/lib/emails/logEmailAttempt";
 
 /**
  * Batch-completion check for the one-digest-per-scrape design (chat#1855):
@@ -20,8 +22,31 @@ export async function maybeSendScrapeDigest(batchId: string | null | undefined)
     .map(r => ({ platform: r.platform ?? "other", postUrls: r.new_post_urls ?? [] }));
   if (!sections.length) return null;
 
-  const emails = await getScrapeDigestRecipients(
+  const { emails, artistIds } = await getScrapeDigestRecipients(
     runs.map(r => r.social_id).filter((id): id is string => Boolean(id)),
   );
-  return await sendScrapeDigestEmail({ emails, sections });
+
+  // Rate cap: at most one digest per watched artist entity per 24h (chat#1855).
+  const since = new Date(Date.now() - 24 * 60 * 60 * 1000).toISOString();
+  const recent = await selectRecentScrapeDigestLogs(artistIds, since);
+  if (recent.length > 0) return null;
+
+  const sent = await sendScrapeDigestEmail({ emails, sections });
+
+  // Audit trail: record the send per artist entity so alerts are debuggable
+  // and the cap has something to read (the solo path never logged at all).
+  const resendId = (sent as { id?: string } | null)?.id;
+  if (sent) {
+    await Promise.all(
+      artistIds.map(artistId =>
+        logEmailAttempt({
+          rawBody: `scrape-digest:${batchId}`,
+          status: "sent",
+          accountId: artistId,
+          resendId,
+        }),
+      ),
+    );
+  }
+  return sent;
 }

lib/supabase/email_send_log/selectRecentScrapeDigestLogs.ts

@@ -0,0 +1,20 @@
+import supabase from "@/lib/supabase/serverClient";
+
+/**
+ * Returns scrape-digest send-log rows for any of the given artist entity ids
+ * newer than `since` — the digest rate cap's lookback (chat#1855).
+ */
+export async function selectRecentScrapeDigestLogs(artistIds: string[], since: string) {
+  if (!artistIds.length) return [];
+  const { data, error } = await supabase
+    .from("email_send_log")
+    .select("account_id, created_at")
+    .in("account_id", artistIds)
+    .like("raw_body", "scrape-digest:%")
+    .gte("created_at", since);
+  if (error) {
+    console.error("[ERROR] selectRecentScrapeDigestLogs:", error);
+    return [];
+  }
+  return data ?? [];
+}