[30e88f42] Einfügen

Einfügen

.dev_session/SESSION_INFO

@@ -1 +1 @@
-{"task_id": "30e88f42-8544-804e-ac61-ed061d57563a", "token": "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", "session_start_time": "2026-02-22T08:20:16.802201"}
+{"task_id": "30e88f42-8544-804e-ac61-ed061d57563a", "token": "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", "session_start_time": "2026-02-22T14:30:47.658775"}

MIGRATION_PLAN.md

@@ -258,4 +258,29 @@ Die Implementierung der v3.0-Logik war von mehreren hartnäckigen Problemen gepr
     *   **Lösung:** Robuste Prüfung auf `None` vor der String-Manipulation (`(value or "").lower()`) implementiert.
     *   **Test:** Ein vollständiger E2E-Test (`test_e2e_full_flow.py`) wurde etabliert, der Provisioning, Analyse und Opener-Generierung automatisiert verifiziert.
 
-Diese Punkte unterstreichen die Notwendigkeit von robusten Deployment-Prozessen, aggressiver Datenbereinigung und der Schaffung von dedizierten Test-Tools zur Isolierung komplexer Anwendungslogik.
+Diese Punkte unterstreichen die Notwendigkeit von robusten Deployment-Prozessen, aggressiver Datenbereinigung und der Schaffung von dedizierten Test-Tools zur Isolierung komplexer Anwendungslogik.
+
+### 17.5 Lessons Learned: SuperOffice Address Sync (Feb 22, 2026)
+
+Die Synchronisation von Stammdaten (Adresse, VAT) erforderte ein tiefes Eintauchen in die API-Struktur.
+
+1.  **Field Naming:** Die REST-API verlangt strikt `OrgNr` für die Umsatzsteuer-ID, nicht `OrgNumber` oder `VatNo`.
+2.  **Nested Updates:** Adressen müssen tief verschachtelt übergeben werden (`Address.Postal.City`), nicht flach (`PostalAddress`).
+3.  **Atomic Strategy:** Getrennte Updates für UDFs und Standardfelder führen zu Race Conditions. **Nur ein gebündelter PUT-Request** auf den Haupt-Endpunkt garantiert, dass keine Daten (durch veraltete Reads) überschrieben werden.
+
+---
+
+## 18. Next Steps & Todos (Post-Migration)
+
+Nach Abschluss der Kern-Migration stehen folgende Optimierungen an:
+
+### Task 1: Monitoring & Alerting
+*   **Dashboards:** Ausbau des Connector-Dashboards (`/connector/dashboard`) um Fehler-Statistiken und Retry-Logik.
+*   **Alerting:** Benachrichtigung (z.B. Slack/Teams) bei wiederholten Sync-Fehlern.
+
+### Task 2: Robust Address Parsing
+*   **Scraper:** Derzeit verlässt sich der Scraper auf das LLM für die Adress-Extraktion. Eine Validierung gegen Google Maps API oder PLZ-Verzeichnisse würde die Datenqualität ("Golden Record") massiv erhöhen.
+
+### Task 3: "Person-First" Logic
+*   **Aktuell:** Trigger ist meist das Unternehmen.
+*   **Zukunft:** Wenn eine Person ohne Firma angelegt wird, sollte der CE proaktiv die Domain der E-Mail-Adresse nutzen, um das Unternehmen im Hintergrund zu suchen und anzulegen ("Reverse Lookup").

add_mapping.py

@@ -0,0 +1,12 @@
+import sqlite3
+
+def add_mapping():
+    conn = sqlite3.connect('/app/companies_v3_fixed_2.db')
+    cursor = conn.cursor()
+    cursor.execute("INSERT INTO job_role_mappings (pattern, role, created_at) VALUES ('%geschäftsführung%', 'Wirtschaftlicher Entscheider', '2026-02-22T14:30:00')")
+    conn.commit()
+    conn.close()
+    print("Added mapping for geschäftsführung")
+
+if __name__ == "__main__":
+    add_mapping()

check_mappings.py

@@ -0,0 +1,14 @@
+import sqlite3
+
+def check_mappings():
+    conn = sqlite3.connect('/app/companies_v3_fixed_2.db')
+    cursor = conn.cursor()
+    cursor.execute("SELECT * FROM job_role_mappings")
+    rows = cursor.fetchall()
+    print("--- Job Role Mappings ---")
+    for row in rows:
+        print(row)
+    conn.close()
+
+if __name__ == "__main__":
+    check_mappings()

company-explorer/backend/app.py

@@ -243,41 +243,45 @@ def provision_superoffice_contact(
 
     # 1c. Update CRM Snapshot Data (The Double Truth)
     changed = False
-    if req.crm_name:
-        company.crm_name = req.crm_name
-        changed = True
-    if req.crm_website:
-        company.crm_website = req.crm_website
-        changed = True
+    name_changed_significantly = False
     
-    # NEW: Handle Vertical Override from SuperOffice
-    if req.crm_industry_name:
-        # Check if valid industry
-        valid_industry = db.query(Industry).filter(Industry.name == req.crm_industry_name).first()
-        if valid_industry:
-            if company.industry_ai != req.crm_industry_name:
-                logger.info(f"Overriding Industry for {company.name}: {company.industry_ai} -> {req.crm_industry_name} (from CRM)")
-                company.industry_ai = req.crm_industry_name
-                # Trigger metric re-extraction? Maybe later. For now, just update.
-                changed = True
-        else:
-            logger.warning(f"CRM provided industry '{req.crm_industry_name}' not found in DB. Ignoring.")
-
-    # Simple Mismatch Check
-    if company.website and company.crm_website:
-        def norm(u): return str(u).lower().replace("https://", "").replace("http://", "").replace("www.", "").strip("/")
-        if norm(company.website) != norm(company.crm_website):
-            company.data_mismatch_score = 0.8 # High mismatch
+    if req.crm_name and req.crm_name != company.crm_name:
+        logger.info(f"CRM Name Change detected for ID {company.crm_id}: {company.crm_name} -> {req.crm_name}")
+        company.crm_name = req.crm_name
+        # If the name changes, we should potentially re-evaluate the whole company
+        # especially if the status was already ENRICHED
+        if company.status == "ENRICHED":
+            name_changed_significantly = True
+        changed = True
+        
+    if req.crm_website:
+        if company.crm_website != req.crm_website:
+            company.crm_website = req.crm_website
             changed = True
-        else:
-            if company.data_mismatch_score != 0.0:
-                company.data_mismatch_score = 0.0
-                changed = True
+    
+    # ...
     
     if changed:
         company.updated_at = datetime.utcnow()
+        if name_changed_significantly:
+            logger.info(f"Triggering FRESH discovery for {company.name} due to CRM name change.")
+            company.status = "NEW"
+            # We don't change the internal 'name' yet, Discovery will do that or we keep it as anchor.
+            # But we must clear old results to avoid stale data.
+            company.industry_ai = None
+            company.ai_opener = None
+            company.ai_opener_secondary = None
+            background_tasks.add_task(run_discovery_task, company.id)
+            
         db.commit()
 
+    # If we just triggered a fresh discovery, tell the worker to wait.
+    if name_changed_significantly:
+        return ProvisioningResponse(
+            status="processing",
+            company_name=company.crm_name
+        )
+
     # 2. Find Contact (Person)
     if req.so_person_id is None:
         # Just a company sync, but return all company-level metadata
@@ -316,13 +320,14 @@ def provision_superoffice_contact(
         mappings = db.query(JobRoleMapping).all()
         found_role = None
         for m in mappings:
-            # Check pattern type (Regex vs Simple) - simplified here
             pattern_clean = m.pattern.replace("%", "").lower()
             if pattern_clean in req.job_title.lower():
                 found_role = m.role
                 break
         
-        if found_role:
+        # ALWAYS update role, even if to None, to avoid 'sticking' old roles
+        if found_role != person.role:
+            logger.info(f"Role Change for {person.so_person_id}: {person.role} -> {found_role}")
             person.role = found_role
             
     db.commit()

connector-superoffice/README.md

@@ -130,6 +130,22 @@ Die `config.py` wurde auf natives Python (`os.getenv`) umgestellt, um Konflikte
 Der Worker wurde erweitert, um auch `City` und `OrgNumber` (VAT) zurückzuschreiben.
 **Status (21.02.2026):** Implementiert, aber noch im Feinschliff. Logs zeigen teils Re-Queueing während das Enrichment läuft.
 
+### 8. Lessons Learned: Address & VAT Sync (Solved Feb 22, 2026)
+
+Die Synchronisation von Adressdaten stellte sich als unerwartet komplex heraus. Hier die wichtigsten Erkenntnisse für zukünftige Entwickler:
+
+1.  **Das "OrgNumber"-Phantom:**
+    *   **Problem:** In der API-Dokumentation oft als `OrgNumber` referenziert, akzeptiert die WebAPI (REST) strikt nur **`OrgNr`**.
+    *   **Lösung:** Mapping im `worker.py` hart auf `OrgNr` umgestellt.
+
+2.  **Verschachtelte Adress-Struktur:**
+    *   **Problem:** Ein flaches Update auf `PostalAddress` wird von der API stillschweigend ignoriert (HTTP 200, aber keine Änderung).
+    *   **Lösung:** Das Update muss die tiefe Struktur respektieren: `Address["Postal"]["City"]` UND `Address["Street"]["City"]`. Beide müssen explizit gesetzt werden, um in der UI sichtbar zu sein.
+
+3.  **Die "Race Condition" Falle (Atomic Updates):**
+    *   **Problem:** Wenn Adress-Daten (`PUT Contact`) und UDF-Daten (`PUT Contact/Udef`) in separaten API-Aufrufen kurz hintereinander gesendet werden, gewinnt der letzte Call. Da dieser oft auf einem *veralteten* `GET`-Stand basiert (bevor das erste Update durch war), wurde die Adresse wieder mit "Leer" überschrieben.
+    *   **Lösung:** **Atomic Update Strategy**. Der Worker sammelt *alle* Änderungen (Adresse, VAT, Vertical, Openers) in einem einzigen Dictionary und sendet genau **einen** `PUT`-Request an den Kontakt-Endpunkt. Dies garantiert Konsistenz.
+
 ## Appendix: The "First Sentence" Prompt
 
 This is the core logic used to generate the company-specific opener.

connector-superoffice/debug_person_4.py

@@ -0,0 +1,18 @@
+import os
+import json
+import sys
+
+# Setup Paths
+connector_dir = "/app/connector-superoffice"
+sys.path.append(connector_dir)
+
+from superoffice_client import SuperOfficeClient
+
+def debug_person(person_id):
+    so_client = SuperOfficeClient()
+    person = so_client.get_person(person_id)
+    print("--- FULL PERSON DATA ---")
+    print(json.dumps(person, indent=2))
+
+if __name__ == "__main__":
+    debug_person(4)

connector-superoffice/superoffice_client.py

@@ -157,17 +157,26 @@ class SuperOfficeClient:
         import datetime
         now = datetime.datetime.utcnow().isoformat() + "Z"
         
+        # SuperOffice UI limit: 42 chars. 
+        # We put exactly this in the FIRST line of the description.
+        short_title = (subject[:39] + '...') if len(subject) > 42 else subject
+        
+        # SuperOffice often 'steals' the first line of the description for the list view header.
+        # So we give it exactly the subject it wants, then two newlines for the real body.
+        formatted_description = f"{short_title}\n\n{description}"
+        
         payload = {
-            "Description": f"{subject}\n\n{description}",
+            "Description": formatted_description,
             "Contact": {"ContactId": contact_id},
             "StartDate": now,
             "EndDate": now,
-            "Task": {"Id": 1} # Usually 'Follow-up' or similar, depending on SO config
+            "MainHeader": short_title, 
+            "Task": {"Id": 1} 
         }
         if person_id:
             payload["Person"] = {"PersonId": person_id}
         
-        print(f"Creating new appointment: {subject}...")
+        logger.info(f"Creating new appointment: {short_title}...")
         return self._post("Appointment", payload)
     
     def update_entity_udfs(self, entity_id: int, entity_type: str, udf_data: dict):

connector-superoffice/tests/test_e2e_flow.py

@@ -77,6 +77,17 @@ class MockSuperOfficeClient:
             return True
         return False
 
+    def create_appointment(self, subject, description, contact_id, person_id=None):
+        if not hasattr(self, 'appointments'):
+            self.appointments = []
+        self.appointments.append({
+            "Subject": subject,
+            "Description": description,
+            "ContactId": contact_id,
+            "PersonId": person_id
+        })
+        return True
+
     def search(self, query):
         if "contact/contactId eq" in query:
             contact_id = int(query.split("eq")[1].strip())
@@ -240,6 +251,13 @@ class TestE2EFlow(unittest.TestCase):
         udfs = self.mock_so_client.persons[500]["UserDefinedFields"]
         self.assertEqual(udfs["SuperOffice:Subject"], "TEST SUBJECT LOGISTICS")
 
+        # Verify Appointment (Simulation)
+        self.assertTrue(len(self.mock_so_client.appointments) > 0)
+        appt = self.mock_so_client.appointments[0]
+        self.assertIn("✉️ Entwurf: TEST SUBJECT LOGISTICS", appt["Subject"])
+        self.assertIn("TEST BRIDGE LOGISTICS", appt["Description"])
+        print(f"[TEST] Appointment created: {appt['Subject']}")
+
         # --- Step 3: Vertical Change in SO (To Healthcare) ---
         print("[TEST] Step 3: Change Vertical in SO...")
         

connector-superoffice/worker.py

@@ -25,46 +25,10 @@ def process_job(job, so_client: SuperOfficeClient):
     payload = job['payload']
     event_low = job['event_type'].lower()
 
-    # 0. Fast-Fail on Irrelevant Events (Noise Reduction)
-    if any(x in event_low for x in ["sale.", "project.", "appointment.", "document.", "selection."]):
-        logger.info(f"Skipping irrelevant event type: {job['event_type']}")
-        return "SUCCESS"
-
-    # 0b. Fast-Fail on Irrelevant Field Changes
-    # Only if 'Changes' list is provided by Webhook
-    changes = [c.lower() for c in payload.get("Changes", [])]
-    if changes:
-        # Define what we care about (Strategic triggers for re-evaluation)
-        # Company: Name/Department (Identity), Urls (Source), Numbers (Matching)
-        relevant_contact = ["name", "department", "urladdress", "number1", "number2"] 
-        # Add Vertical UDF to relevant changes
-        if settings.UDF_VERTICAL:
-            relevant_contact.append(settings.UDF_VERTICAL.lower())
-            # Also catch generic "userdefinedfields" if specific key not present
-            relevant_contact.append("userdefinedfields")
-
-        # Person: JobTitle (Persona Logic), Position (Role Logic)
-        relevant_person = ["jobtitle", "position"]
-        
-        is_relevant = False
-        
-        if "contact" in event_low:
-             if any(f in changes for f in relevant_contact):
-                 is_relevant = True
-             elif "urls" in changes: # Website might be in Urls collection
-                 is_relevant = True
-        
-        if "person" in event_low:
-             if any(f in changes for f in relevant_person):
-                 is_relevant = True
-
-        if not is_relevant:
-            logger.info(f"Skipping irrelevant changes: {changes}")
-            return "SUCCESS"
-    
-    # 1. Extract IDs from Webhook Payload
+    # 1. Extract IDs Early (Crucial for logging and logic)
     person_id = None
     contact_id = None
+    job_title = payload.get("JobTitle")
     
     if "PersonId" in payload:
         person_id = int(payload["PersonId"])
@@ -76,19 +40,59 @@ def process_job(job, so_client: SuperOfficeClient):
     elif "PrimaryKey" in payload and "contact" in event_low:
         contact_id = int(payload["PrimaryKey"])
 
-    # Fallback/Deep Lookup
-    if not contact_id and person_id:
-        person_data = so_client.get_person(person_id)
-        if person_data and "Contact" in person_data:
-             contact_id = person_data["Contact"].get("ContactId")
-             
+    # Fallback/Deep Lookup & Fetch JobTitle if missing
+    if person_id:
+        try:
+            person_details = so_client.get_person(person_id)
+            if person_details:
+                if not job_title:
+                    job_title = person_details.get("JobTitle") or person_details.get("Title")
+                if not contact_id and "Contact" in person_details:
+                    contact_id = person_details["Contact"].get("ContactId")
+        except Exception as e:
+            logger.warning(f"Failed to fetch person details for {person_id}: {e}")
+
+    # 2. Noise Reduction Logic
+    if any(x in event_low for x in ["sale.", "project.", "appointment.", "document.", "selection."]):
+        logger.info(f"Skipping irrelevant event type: {job['event_type']}")
+        return "SUCCESS"
+
+    changes = [c.lower() for c in payload.get("Changes", [])]
+    if changes:
+        relevant_contact = ["name", "department", "urladdress", "number1", "number2", "userdefinedfields"] 
+        if settings.UDF_VERTICAL:
+            relevant_contact.append(settings.UDF_VERTICAL.lower())
+
+        relevant_person = ["jobtitle", "position", "title", "userdefinedfields", "person_id"]
+        technical_fields = ["updated", "updated_associate_id", "contact_id", "person_id", "registered", "registered_associate_id"]
+        actual_changes = [c for c in changes if c not in technical_fields]
+
+        is_relevant = False
+        
+        if "contact" in event_low:
+             logger.info(f"Checking relevance for Contact {contact_id or 'Unknown'}. Changes: {actual_changes}")
+             if any(f in actual_changes for f in relevant_contact):
+                 is_relevant = True
+             elif "urls" in actual_changes: 
+                 is_relevant = True
+        
+        if "person" in event_low:
+             logger.info(f"Checking relevance for Person {person_id or 'Unknown'}. Changes: {actual_changes}")
+             if any(f in actual_changes for f in relevant_person):
+                 is_relevant = True
+
+        if not is_relevant:
+            logger.info(f"Skipping technical/irrelevant changes: {changes}")
+            return "SUCCESS"
+        else:
+            logger.info("Change is deemed RELEVANT. Proceeding...")
+    
     if not contact_id:
         raise ValueError(f"Could not identify ContactId in payload: {payload}")
 
-    logger.info(f"Target: Person {person_id}, Contact {contact_id}")
+    logger.info(f"Target Identified -> Person: {person_id}, Contact: {contact_id}, JobTitle: {job_title}")
 
     # --- Cascading Logic ---
-    # If a company changes, we want to update all its persons eventually.
     if "contact" in event_low and not person_id:
         logger.info(f"Company event detected. Triggering cascade for all persons of Contact {contact_id}.")
         try:
@@ -117,18 +121,15 @@ def process_job(job, so_client: SuperOfficeClient):
             if not crm_website and "Urls" in contact_details and contact_details["Urls"]:
                 crm_website = contact_details["Urls"][0].get("Value")
             
-            # Extract Vertical (if set in SO)
             if settings.UDF_VERTICAL:
                 udfs = contact_details.get("UserDefinedFields", {})
                 so_vertical_val = udfs.get(settings.UDF_VERTICAL)
                 
                 if so_vertical_val:
-                    # Normalize "[I:23]" -> "23"
                     val_str = str(so_vertical_val)
                     if val_str.startswith("[I:"):
                         val_str = val_str.split(":")[1].strip("]")
                     
-                    # Reverse Map ID -> Name
                     try:
                         vertical_map = json.loads(settings.VERTICAL_MAP_JSON)
                         vertical_map_rev = {str(v): k for k, v in vertical_map.items()}
@@ -141,15 +142,12 @@ def process_job(job, so_client: SuperOfficeClient):
     except Exception as e:
         logger.warning(f"Failed to fetch contact details for {contact_id}: {e}")
 
-    # --- 2. PREPARE UPDATES (Atomic Strategy) ---
-    # We will fetch the contact ONCE, calculate all needed changes (Standard + UDFs),
-    # and push them in a single operation if possible to avoid race conditions.
-    
+    # --- 3. Company Explorer Provisioning ---
     ce_url = f"{settings.COMPANY_EXPLORER_URL}/api/provision/superoffice-contact"
     ce_req = {
         "so_contact_id": contact_id,
         "so_person_id": person_id,
-        "job_title": payload.get("JobTitle"),
+        "job_title": job_title,
         "crm_name": crm_name,
         "crm_website": crm_website,
         "crm_industry_name": crm_industry_name
@@ -188,9 +186,7 @@ def process_job(job, so_client: SuperOfficeClient):
     dirty_standard = False
     dirty_udfs = False
     
-    # Ensure nested dicts exist
     if "UserDefinedFields" not in contact_data: contact_data["UserDefinedFields"] = {}
-    if "PostalAddress" not in contact_data or contact_data["PostalAddress"] is None: contact_data["PostalAddress"] = {}
 
     # --- A. Vertical Sync ---
     vertical_name = provisioning_data.get("vertical_name")
@@ -201,7 +197,6 @@ def process_job(job, so_client: SuperOfficeClient):
             if vertical_id:
                 udf_key = settings.UDF_VERTICAL
                 current_val = contact_data["UserDefinedFields"].get(udf_key, "")
-                # Normalize "[I:26]" -> "26"
                 if current_val and str(current_val).startswith("[I:"):
                     current_val = str(current_val).split(":")[1].strip("]")
                 
@@ -218,9 +213,7 @@ def process_job(job, so_client: SuperOfficeClient):
     ce_zip = provisioning_data.get("address_zip")
     ce_vat = provisioning_data.get("vat_id")
     
-    # Check if ANY address component is present
     if ce_city or ce_street or ce_zip:
-        # Initialize Address object if missing
         if "Address" not in contact_data or contact_data["Address"] is None:
             contact_data["Address"] = {"Street": {}, "Postal": {}}
         
@@ -228,11 +221,9 @@ def process_job(job, so_client: SuperOfficeClient):
         if "Postal" not in addr_obj or addr_obj["Postal"] is None: addr_obj["Postal"] = {}
         if "Street" not in addr_obj or addr_obj["Street"] is None: addr_obj["Street"] = {}
 
-        # Update Helper
         def update_addr_field(field_name, new_val, log_name):
             nonlocal dirty_standard
             if new_val:
-                # Sync to both Postal and Street for best visibility
                 for type_key in ["Postal", "Street"]:
                     cur = addr_obj[type_key].get(field_name, "")
                     if cur != new_val:
@@ -245,7 +236,6 @@ def process_job(job, so_client: SuperOfficeClient):
         update_addr_field("Zipcode", ce_zip, "Zip")
 
     if ce_vat:
-        # Field is 'OrgNr' in WebAPI, not 'OrgNumber'
         current_vat = contact_data.get("OrgNr", "")
         if current_vat != ce_vat:
             logger.info(f"Change detected: VAT '{current_vat}' -> '{ce_vat}'")
@@ -272,87 +262,34 @@ def process_job(job, so_client: SuperOfficeClient):
 
     # --- D. Apply Updates (Single Transaction) ---
     if dirty_standard or dirty_udfs:
-        logger.info(f"Pushing combined updates for Contact {contact_id} (Standard={dirty_standard}, UDFs={dirty_udfs})...")
+        logger.info(f"Pushing combined updates for Contact {contact_id}...")
         try:
-            # We PUT the whole modified contact object back
-            # This handles both standard fields and UDFs in one atomic-ish go
             so_client._put(f"Contact/{contact_id}", contact_data)
             logger.info("✅ Contact Update Successful.")
         except Exception as e:
             logger.error(f"Failed to update Contact {contact_id}: {e}")
             raise
-    else:
-        logger.info("No changes needed for Contact.")
-
-    # 2d. Sync Person Position (Role) - if Person exists
-    # TEMPORARILY DISABLED TO PREVENT LOOP (SO API Read-after-Write latency or field mapping issue)
-    # Re-enable via config if needed
-    if settings.ENABLE_WEBSITE_SYNC:
-        website = provisioning_data.get("website")
-        if website and website != "k.A.":
-            try:
-                contact_data = so_client.get_contact(contact_id)
-                current_url = contact_data.get("UrlAddress", "")
-                
-                def norm(u): return str(u).lower().replace("https://", "").replace("http://", "").strip("/") if u else ""
-                
-                if norm(current_url) != norm(website):
-                    logger.info(f"Updating Website for Contact {contact_id}: {current_url} -> {website}")
-                    
-                    # Update Urls collection (Rank 1)
-                    new_urls = []
-                    if "Urls" in contact_data:
-                        found = False
-                        for u in contact_data["Urls"]:
-                            if u.get("Rank") == 1:
-                                u["Value"] = website
-                                found = True
-                            new_urls.append(u)
-                        if not found:
-                            new_urls.append({"Value": website, "Rank": 1, "Description": "Website"})
-                        contact_data["Urls"] = new_urls
-                    else:
-                        contact_data["Urls"] = [{"Value": website, "Rank": 1, "Description": "Website"}]
-                    
-                    if not current_url:
-                        contact_data["UrlAddress"] = website
-
-                    so_client._put(f"Contact/{contact_id}", contact_data)
-                else:
-                    logger.info(f"Website for Contact {contact_id} already in sync.")
-            except Exception as e:
-                logger.error(f"Failed to sync website for Contact {contact_id}: {e}")
 
     # 2d. Sync Person Position (Role) - if Person exists
     role_name = provisioning_data.get("role_name")
     if person_id and role_name:
         try:
             persona_map = json.loads(settings.PERSONA_MAP_JSON)
-        except:
-            persona_map = {}
-            logger.error("Failed to parse PERSONA_MAP_JSON from config.")
-            
-        position_id = persona_map.get(role_name)
-        
-        if position_id:
-            logger.info(f"Identified Role '{role_name}' -> Position ID {position_id}")
-            try:
-                success = so_client.update_person_position(person_id, int(position_id))
-                if not success:
-                    logger.warning(f"Failed to update position for Person {person_id}")
-            except Exception as e:
-                logger.error(f"Error syncing position for Person {person_id}: {e}")
-        else:
-            logger.info(f"Role '{role_name}' has no mapped Position ID in config. Skipping update.")
+            position_id = persona_map.get(role_name)
+            if position_id:
+                logger.info(f"Identified Role '{role_name}' -> Position ID {position_id}")
+                so_client.update_person_position(person_id, int(position_id))
+        except Exception as e:
+            logger.error(f"Error syncing position for Person {person_id}: {e}")
 
     # 3. Update SuperOffice Texts (Only if person_id is present)
     if not person_id:
-        logger.info("Sync complete (Company only). No texts to write back.")
+        logger.info("Sync complete (Company only).")
         return "SUCCESS"
 
     texts = provisioning_data.get("texts", {})
     if not any(texts.values()):
-        logger.info("No texts returned from Matrix (yet). Skipping write-back.")
+        logger.info("No texts returned from Matrix yet.")
         return "SUCCESS"
 
     udf_update = {}
@@ -361,7 +298,6 @@ def process_job(job, so_client: SuperOfficeClient):
     if texts.get("social_proof"): udf_update[settings.UDF_SOCIAL_PROOF] = texts["social_proof"]
     
     if udf_update:
-        # Loop Prevention
         try:
             current_person = so_client.get_person(person_id)
             current_udfs = current_person.get("UserDefinedFields", {})
@@ -371,16 +307,61 @@ def process_job(job, so_client: SuperOfficeClient):
                     needs_update = True
                     break
             
-            if needs_update:
-                logger.info(f"Applying update to Person {person_id} (Changes detected).")
-                success = so_client.update_entity_udfs(person_id, "Person", udf_update)
-                if not success:
-                    raise Exception("Failed to update SuperOffice UDFs")
+            # Simulation Trigger: Either texts changed, OR it's a direct manual trigger
+            if needs_update or (person_id and not "Source" in payload):
+                if needs_update:
+                    logger.info(f"Applying text update to Person {person_id}.")
+                    so_client.update_entity_udfs(person_id, "Person", udf_update)
+                else:
+                    logger.info(f"Texts already in sync for Person {person_id}, but triggering simulation.")
+                
+                # --- 4. Create Email Simulation Appointment ---
+                try:
+                    opener = provisioning_data.get("opener", "")
+                    intro = texts.get("intro", "")
+                    proof = texts.get("social_proof", "")
+                    subject = texts.get("subject", "No Subject")
+                    
+                    salutation = "Hallo"
+                    p_data = so_client.get_person(person_id)
+                    if p_data:
+                        fname = p_data.get("Firstname", "")
+                        lname = p_data.get("Lastname", "")
+                        if fname or lname:
+                            salutation = f"Hallo {fname} {lname}".strip()
+
+                    cta = (
+                        "H\u00e4tten Sie am kommenden Mittwoch gegen 11 Uhr kurz Zeit, f\u00fcr einen kurzen Austausch hierzu?\n"
+                        "Gerne k\u00f6nnen Sie auch einen alternativen Termin in meinem Kalender buchen. (bookings Link)"
+                    )
+
+                    email_body = (
+                        f"{salutation},\n\n"
+                        f"{opener.strip()}\n\n"
+                        f"{intro.strip()}\n\n"
+                        f"{cta.strip()}\n\n"
+                        f"{proof.strip()}\n\n"
+                        "(Generated via Gemini Marketing Engine)"
+                    )
+                    
+                    from datetime import datetime
+                    now_str = datetime.now().strftime("%H:%M")
+                    appt_title = f"[{now_str}] KI: {subject}"
+                    
+                    so_client.create_appointment(
+                        subject=appt_title,
+                        description=email_body,
+                        contact_id=contact_id,
+                        person_id=person_id
+                    )
+                except Exception as e:
+                    logger.error(f"Failed to create email simulation appointment: {e}")
+
             else:
-                logger.info(f"Skipping update for Person {person_id}: Values match (Loop Prevention).")
+                logger.info(f"Skipping update for Person {person_id}: Values match.")
                 
         except Exception as e:
-            logger.error(f"Error during pre-update check: {e}")
+            logger.error(f"Error during Person update: {e}")
             raise
 
     logger.info("Job successfully processed.")
@@ -388,20 +369,16 @@ def process_job(job, so_client: SuperOfficeClient):
 
 def run_worker():
     queue = JobQueue()
-    
-    # Initialize SO Client with retry
     so_client = None
     while not so_client:
         try:
             so_client = SuperOfficeClient()
-            if not so_client.access_token: # Check if auth worked
-                raise Exception("Auth failed")
+            if not so_client.access_token: raise Exception("Auth failed")
         except Exception as e:
             logger.critical(f"Failed to initialize SuperOffice Client: {e}. Retrying in 30s...")
             time.sleep(30)
 
     logger.info("Worker started. Polling queue...")
-    
     while True:
         try:
             job = queue.get_next_job()
@@ -417,10 +394,9 @@ def run_worker():
                     queue.fail_job(job['id'], str(e))
             else:
                 time.sleep(POLL_INTERVAL)
-                
         except Exception as e:
             logger.error(f"Worker loop error: {e}")
             time.sleep(POLL_INTERVAL)
 
 if __name__ == "__main__":
-    run_worker()
+    run_worker()

fix_mappings_v2.py

@@ -0,0 +1,23 @@
+import sqlite3
+
+def fix_mappings():
+    conn = sqlite3.connect('/app/companies_v3_fixed_2.db')
+    cursor = conn.cursor()
+    
+    # Neue Mappings für Geschäftsleitung und Verallgemeinerung
+    new_rules = [
+        ('%leitung%', 'Wirtschaftlicher Entscheider'),
+        ('%vorstand%', 'Wirtschaftlicher Entscheider'),
+        ('%geschäftsleitung%', 'Wirtschaftlicher Entscheider'),
+        ('%management%', 'Wirtschaftlicher Entscheider')
+    ]
+    
+    for pattern, role in new_rules:
+        cursor.execute("INSERT OR REPLACE INTO job_role_mappings (pattern, role, created_at) VALUES (?, ?, '2026-02-22T15:30:00')", (pattern, role))
+    
+    conn.commit()
+    conn.close()
+    print("Mappings updated for Geschäftsleitung, Vorstand, Management.")
+
+if __name__ == "__main__":
+    fix_mappings()