Import ClientFlow production v4928.1.5.132.4

scripts/audit_contact_identity_collisions.py

@@ -0,0 +1,78 @@
+#!/usr/bin/env python3
+"""Audit possible polluted Chatwoot contact names in pending work items.
+
+Read-only. It highlights sender names reused across several different sender
+emails, which can make the Operator Workbench display the wrong person name.
+"""
+from __future__ import annotations
+
+import re
+from collections import defaultdict
+from sqlalchemy import text
+
+from app.db import engine
+
+
+def norm(value: object) -> str:
+    return " ".join(re.sub(r"[^0-9a-zA-ZÀ-ÿ]+", " ", str(value or "").casefold()).split())
+
+
+SQL = """
+    SELECT
+        t.id::text AS task_id,
+        t.created_at,
+        t.status,
+        t.action_code,
+        t.conversation_id,
+        COALESCE(re.payload->'sender'->>'name', '') AS sender_name,
+        COALESCE(re.payload->'sender'->>'email', '') AS sender_email,
+        COALESCE(o.title, '') AS opportunity_title,
+        COALESCE(cu.name, '') AS fiscal_customer_name
+    FROM tasks t
+    LEFT JOIN raw_events re ON re.id = t.raw_event_id
+    LEFT JOIN opportunities o ON o.id = t.opportunity_id
+    LEFT JOIN customers cu ON cu.id = o.local_customer_id
+    WHERE t.status = 'pending'
+      AND t.source_system = 'chatwoot'
+    ORDER BY t.created_at DESC
+    LIMIT :limit
+"""
+
+
+def main() -> int:
+    with engine.begin() as conn:
+        rows = [dict(r) for r in conn.execute(text(SQL), {"limit": 250}).mappings().all()]
+
+    by_name: dict[str, list[dict]] = defaultdict(list)
+    for row in rows:
+        key = norm(row.get("sender_name"))
+        if key:
+            by_name[key].append(row)
+
+    suspects = []
+    for key, items in by_name.items():
+        emails = {str(i.get("sender_email") or "").casefold() for i in items if i.get("sender_email")}
+        if len(emails) > 1:
+            suspects.append((key, emails, items))
+
+    print("Contact identity collision audit")
+    print("=" * 80)
+    print(f"pending_chatwoot_tasks={len(rows)}")
+    print(f"suspect_reused_names={len(suspects)}")
+    print()
+
+    for key, emails, items in sorted(suspects, key=lambda x: (-len(x[1]), x[0])):
+        display = items[0].get("sender_name") or key
+        print("=" * 80)
+        print(f"sender_name={display!r} reused_with_{len(emails)}_emails")
+        for item in items[:20]:
+            print(
+                f"- task={item['task_id']} conv={item.get('conversation_id')} "
+                f"email={item.get('sender_email') or '-'} action={item.get('action_code')} "
+                f"opportunity={item.get('opportunity_title') or '-'} fiscal={item.get('fiscal_customer_name') or '-'}"
+            )
+    return 1 if suspects else 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())