email: fix backfill crash on emails with no Reply-To; Sync now retries errored mailboxes (v0.1.0:62)

insert_email's recipients loop did `for a in parsed.get(kind, [])`, but the parser sets
reply_to=None when there is no Reply-To header, so .get returns None (key present) and the
loop raised 'NoneType' object is not iterable — aborting the entire Gmail backfill on the
first such email (i.e. almost immediately). Fixed with `or []`. Regression test
test_insert_email.py (reply_to=None, all-None recipients, happy path).

Because the scheduler intentionally skips error-status accounts (no retry storms), an
errored mailbox would never resume on its own. "Sync now" now clears error status first,
so it is an explicit retry; backfill resumes from its saved cursor and dedups by
Message-ID, so nothing is re-captured.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

backend/email_integration/db.py

@@ -188,7 +188,10 @@ def insert_email(conn: sqlite3.Connection, *, parsed: dict, match_status: str) -
         elif kind == "reply_to" and parsed.get("reply_to"):
             addrs = [(parsed["reply_to"], None)]
         else:
-            for a in parsed.get(kind, []):
+            # `or []` (not get(kind, [])): the key is often present with value None
+            # (e.g. reply_to when there is no Reply-To header), and `for a in None`
+            # would raise TypeError and abort the whole backfill.
+            for a in (parsed.get(kind) or []):
                 if isinstance(a, dict):
                     addrs.append((a.get("email"), a.get("name")))
                 else:

backend/email_integration/routes.py

@@ -377,6 +377,17 @@ def _h_run_now(handler):
     # Reuse existing rate limit so admins can't hammer this.
     if handler.rate_limited("email-sync-now", 6):
         return handler.send_error_json("Too many requests", 429)
+    # A manual sync is an explicit retry. The scheduler intentionally skips
+    # error-status accounts (no retry storms), so clear that status here so a
+    # mailbox that previously errored is re-attempted. Backfill resumes from its
+    # saved cursor and dedups by Message-ID, so nothing is re-captured twice.
+    conn = _conn()
+    try:
+        conn.execute("UPDATE email_accounts SET sync_status = 'pending', sync_error = NULL "
+                     "WHERE sync_enabled = 1 AND sync_status = 'error'")
+        conn.commit()
+    finally:
+        conn.close()
     result = _sched.trigger_run_now()
     handler.send_json(result)
 

backend/email_integration/test_insert_email.py

@@ -0,0 +1,86 @@
+#!/usr/bin/env python3
+"""Regression test for insert_email: a parsed email with no Reply-To header (reply_to=None)
+must not crash the recipients loop. This bug (`for a in parsed.get('reply_to', [])` returning
+None because the key is present with value None) aborted the whole Gmail backfill on the first
+email lacking a Reply-To header. Synthetic data only (guardrail #9).
+Run: cd backend && python3 email_integration/test_insert_email.py
+"""
+import os
+import sqlite3
+import sys
+
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+from email_integration import db as edb  # noqa: E402
+
+FAILS = []
+
+
+def check(cond, msg):
+    print(("  PASS " if cond else "  FAIL ") + msg)
+    if not cond:
+        FAILS.append(msg)
+
+
+def fresh_conn():
+    conn = sqlite3.connect(":memory:")
+    conn.row_factory = sqlite3.Row
+    edb.apply_migrations(conn.cursor())
+    return conn
+
+
+def main():
+    # 1) the exact crash case: no Reply-To header -> reply_to is None
+    conn = fresh_conn()
+    parsed = {
+        "rfc_message_id": "<m1@example.com>",
+        "from_email": "lp@example.com", "from_name": "An LP",
+        "sent_at": "2026-05-01T10:00:00Z", "subject": "Re: the fund",
+        "to": [{"email": "grant@ten31.xyz", "name": "Grant"}],
+        "cc": [], "bcc": [], "references": [],
+        "reply_to": None,            # <-- previously crashed: 'NoneType' object is not iterable
+        "body_text": "Some concern about lock-up.",
+    }
+    try:
+        eid = edb.insert_email(conn, parsed=parsed, match_status="unmatched")
+        ok = bool(eid)
+    except TypeError as e:
+        ok = False
+        print("  (raised)", e)
+    check(ok, "insert_email with reply_to=None does not raise")
+    if ok:
+        kinds = sorted(r["kind"] for r in conn.execute("SELECT kind FROM email_recipients WHERE email_id=?", (eid,)))
+        check(kinds == ["from", "to"], f"recipients are from+to, reply_to skipped (got {kinds})")
+
+    # 2) defensive: every address field present-but-None must not crash either
+    conn2 = fresh_conn()
+    parsed2 = {
+        "rfc_message_id": "<m2@example.com>", "from_email": "x@example.com",
+        "sent_at": "2026-05-02T10:00:00Z",
+        "to": None, "cc": None, "bcc": None, "references": None, "reply_to": None,
+        "body_text": "no recipients parsed",
+    }
+    try:
+        eid2 = edb.insert_email(conn2, parsed=parsed2, match_status="unmatched")
+        ok2 = bool(eid2)
+    except TypeError as e:
+        ok2 = False
+        print("  (raised)", e)
+    check(ok2, "insert_email with all recipient fields None does not raise")
+
+    # 3) the happy path still records a real Reply-To
+    conn3 = fresh_conn()
+    parsed3 = dict(parsed, rfc_message_id="<m3@example.com>", reply_to="replies@example.com")
+    eid3 = edb.insert_email(conn3, parsed=parsed3, match_status="matched")
+    rt = conn3.execute("SELECT address FROM email_recipients WHERE email_id=? AND kind='reply_to'", (eid3,)).fetchone()
+    check(rt and rt["address"] == "replies@example.com", "a present Reply-To is still recorded")
+
+    if FAILS:
+        print(f"\nFAILED ({len(FAILS)})")
+        for f in FAILS:
+            print("  - " + f)
+        sys.exit(1)
+    print("\nALL PASS (insert_email reply_to/None regression)")
+
+
+if __name__ == "__main__":
+    main()

start9/0.4/startos/utils.ts

@@ -26,8 +26,9 @@ export const PACKAGE_TITLE = 'Ten31 Database'
 //   * 0.1.0:58 (seed 5 Architect positioning framings into the Workshop as candidate options)
 //   * 0.1.0:59 (Email Capture admin panel + matched email into the grounding corpus)
 //   * 0.1.0:60 (Email Capture: single-mailbox enroll field for testing)
-//   * Current:        0.1.0:61 (Email Capture: live backfill progress + auto-refresh)
-export const PACKAGE_VERSION = '0.1.0:61'
+//   * 0.1.0:61 (Email Capture: live backfill progress + auto-refresh)
+//   * Current:        0.1.0:62 (fix backfill crash on no-Reply-To emails; Sync now retries errored mailboxes)
+export const PACKAGE_VERSION = '0.1.0:62'
 
 export const DATA_MOUNT_PATH = '/data'
 export const WEB_PORT = 8080

start9/0.4/startos/versions/index.ts

@@ -22,8 +22,9 @@ import { v_0_1_0_58 } from './v0.1.0.58'
 import { v_0_1_0_59 } from './v0.1.0.59'
 import { v_0_1_0_60 } from './v0.1.0.60'
 import { v_0_1_0_61 } from './v0.1.0.61'
+import { v_0_1_0_62 } from './v0.1.0.62'
 
 export const versionGraph = VersionGraph.of({
-  current: v_0_1_0_61,
-  other: [v_0_1_0_39, v_0_1_0_40, v_0_1_0_41, v_0_1_0_42, v_0_1_0_43, v_0_1_0_44, v_0_1_0_45, v_0_1_0_46, v_0_1_0_47, v_0_1_0_48, v_0_1_0_49, v_0_1_0_50, v_0_1_0_51, v_0_1_0_52, v_0_1_0_53, v_0_1_0_54, v_0_1_0_55, v_0_1_0_56, v_0_1_0_57, v_0_1_0_58, v_0_1_0_59, v_0_1_0_60],
+  current: v_0_1_0_62,
+  other: [v_0_1_0_39, v_0_1_0_40, v_0_1_0_41, v_0_1_0_42, v_0_1_0_43, v_0_1_0_44, v_0_1_0_45, v_0_1_0_46, v_0_1_0_47, v_0_1_0_48, v_0_1_0_49, v_0_1_0_50, v_0_1_0_51, v_0_1_0_52, v_0_1_0_53, v_0_1_0_54, v_0_1_0_55, v_0_1_0_56, v_0_1_0_57, v_0_1_0_58, v_0_1_0_59, v_0_1_0_60, v_0_1_0_61],
 })

start9/0.4/startos/versions/v0.1.0.62.ts

@@ -0,0 +1,19 @@
+import { VersionInfo } from '@start9labs/start-sdk'
+
+// Bugfix: the Gmail backfill crashed with "'NoneType' object is not iterable" on any
+// email with no Reply-To header (parsed reply_to=None hit `for a in parsed.get(kind, [])`,
+// which returns None when the key is present-but-None) — aborting the whole backfill.
+// Fixed with `or []`. Also: "Sync now" now clears error status so a previously-errored
+// mailbox is retried (the scheduler deliberately skips error accounts); backfill resumes
+// from its saved cursor and dedups by Message-ID. No schema migration.
+export const v_0_1_0_62 = VersionInfo.of({
+  version: '0.1.0:62',
+  releaseNotes: {
+    en_US: [
+      'Fixes a bug that stopped the Gmail backfill on any email without a Reply-To header',
+      '(most of them), which is why capture stalled. After updating, open Email Capture and',
+      'click Sync now to resume — it picks up where it left off and will not double-capture.',
+    ].join(' '),
+  },
+  migrations: { up: async () => {}, down: async () => {} },
+})