Floke/Brancheneinstufung2
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: Enhanced CE schema and Notion sync (Pains/Gains)

Browse Source
This commit is contained in:
Moltbot-Jarvis
2026-02-17 19:44:42 +00:00
parent d2e9ee2a70
commit 42bbcd1425
16 changed files with 1731 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

28
company-explorer/backend/database.py
View File

@@ -18,21 +18,35 @@ class Company(Base):
id = Column(Integer, primary_key=True, index=True)
# Core Identity
# Core Identity (Golden Record - from Research)
name = Column(String, index=True)
website = Column(String, index=True) # Normalized Domain preferred
crm_id = Column(String, unique=True, index=True, nullable=True) # Link to D365
# CRM Original Data (Source of Truth for Import)
crm_name = Column(String, nullable=True)
crm_website = Column(String, nullable=True)
crm_address = Column(String, nullable=True) # Full address string or JSON
crm_vat = Column(String, nullable=True)
# Classification
industry_crm = Column(String, nullable=True) # The "allowed" industry
industry_ai = Column(String, nullable=True) # The AI suggested industry
# Location
# Location (Golden Record)
city = Column(String, nullable=True)
country = Column(String, default="DE")
# Workflow Status
status = Column(String, default="NEW", index=True)
status = Column(String, default="NEW", index=True) # NEW, TO_ENRICH, ENRICHED, QUALIFIED, DISQUALIFIED
# Quality & Confidence
confidence_score = Column(Float, default=0.0) # Overall confidence
data_mismatch_score = Column(Float, default=0.0) # 0.0=Match, 1.0=Mismatch
# Scraping Status Flags
website_scrape_status = Column(String, default="PENDING") # PENDING, SUCCESS, FAILED, BLOCKED
wiki_search_status = Column(String, default="PENDING") # PENDING, FOUND, NOT_FOUND
# Granular Process Tracking (Timestamps)
created_at = Column(DateTime, default=datetime.utcnow)
@@ -106,6 +120,13 @@ class Industry(Base):
status_notion = Column(String, nullable=True) # e.g. "P1 Focus Industry"
is_focus = Column(Boolean, default=False) # Derived from status_notion
# Enhanced Fields (v3.1 - Pains/Gains/Priority)
pains = Column(Text, nullable=True)
gains = Column(Text, nullable=True)
notes = Column(Text, nullable=True)
priority = Column(String, nullable=True) # Replaces old status concept ("Freigegeben")
ops_focus_secondary = Column(Boolean, default=False)
# NEW SCHEMA FIELDS (from MIGRATION_PLAN)
metric_type = Column(String, nullable=True) # Unit_Count, Area_in, Area_out
min_requirement = Column(Float, nullable=True)
@@ -117,6 +138,7 @@ class Industry(Base):
# Optional link to a Robotics Category (the "product" relevant for this industry)
primary_category_id = Column(Integer, ForeignKey("robotics_categories.id"), nullable=True)
secondary_category_id = Column(Integer, ForeignKey("robotics_categories.id"), nullable=True)
created_at = Column(DateTime, default=datetime.utcnow)

53
company-explorer/backend/scripts/migrate_ce_db.py Normal file
View File

@@ -0,0 +1,53 @@
import sqlite3
import os
# Adjust path to your actual DB location
DB_PATH = "/home/node/clawd/repos/brancheneinstufung2/company_explorer.db"
def migrate():
if not os.path.exists(DB_PATH):
print(f"Database not found at {DB_PATH}. Maybe it hasn't been created yet?")
return
print(f"Migrating database at {DB_PATH}...")
conn = sqlite3.connect(DB_PATH)
cursor = conn.cursor()
columns_to_add = [
# Industries (Existing List)
("industries", "pains", "TEXT"),
("industries", "gains", "TEXT"),
("industries", "notes", "TEXT"),
("industries", "priority", "TEXT"),
("industries", "ops_focus_secondary", "BOOLEAN DEFAULT 0"),
("industries", "secondary_category_id", "INTEGER"),
# Companies (New List for CRM Data)
("companies", "crm_name", "TEXT"),
("companies", "crm_website", "TEXT"),
("companies", "crm_address", "TEXT"),
("companies", "crm_vat", "TEXT"),
# Companies (Status & Quality)
("companies", "confidence_score", "FLOAT DEFAULT 0.0"),
("companies", "data_mismatch_score", "FLOAT DEFAULT 0.0"),
("companies", "website_scrape_status", "TEXT DEFAULT 'PENDING'"),
("companies", "wiki_search_status", "TEXT DEFAULT 'PENDING'"),
]
for table, col_name, col_type in columns_to_add:
try:
print(f"Adding column '{col_name}' to '{table}'...")
cursor.execute(f"ALTER TABLE {table} ADD COLUMN {col_name} {col_type}")
except sqlite3.OperationalError as e:
if "duplicate column name" in str(e):
print(f"Column '{col_name}' already exists. Skipping.")
else:
print(f"Error adding '{col_name}' to '{table}': {e}")
conn.commit()
conn.close()
print("Migration complete.")
if __name__ == "__main__":
migrate()

170
company-explorer/backend/scripts/sync_notion_to_ce_enhanced.py Normal file
View File

@@ -0,0 +1,170 @@
import sys
import os
import requests
import logging
# Setup Paths
sys.path.append(os.path.abspath("/home/node/clawd/repos/brancheneinstufung2/company-explorer"))
sys.path.append(os.path.abspath("/home/node/clawd/repos/brancheneinstufung2"))
from backend.database import SessionLocal, Industry, RoboticsCategory, init_db
from dotenv import load_dotenv
load_dotenv(dotenv_path="/home/node/clawd/.env")
# Logging
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__)
NOTION_TOKEN = os.getenv("NOTION_API_KEY")
if not NOTION_TOKEN:
logger.error("NOTION_API_KEY missing!")
sys.exit(1)
HEADERS = {
"Authorization": f"Bearer {NOTION_TOKEN}",
"Notion-Version": "2022-06-28",
"Content-Type": "application/json"
}
def find_db_id(query_name):
url = "https://api.notion.com/v1/search"
payload = {"query": query_name, "filter": {"value": "database", "property": "object"}}
resp = requests.post(url, headers=HEADERS, json=payload)
if resp.status_code == 200:
results = resp.json().get("results", [])
if results:
return results[0]['id']
return None
def query_all(db_id):
url = f"https://api.notion.com/v1/databases/{db_id}/query"
results = []
has_more = True
next_cursor = None
while has_more:
payload = {}
if next_cursor: payload["start_cursor"] = next_cursor
resp = requests.post(url, headers=HEADERS, json=payload)
data = resp.json()
results.extend(data.get("results", []))
has_more = data.get("has_more", False)
next_cursor = data.get("next_cursor")
return results
def extract_rich_text(prop):
if not prop or "rich_text" not in prop: return ""
return "".join([t.get("plain_text", "") for t in prop.get("rich_text", [])])
def extract_title(prop):
if not prop or "title" not in prop: return ""
return "".join([t.get("plain_text", "") for t in prop.get("title", [])])
def extract_select(prop):
if not prop or "select" not in prop or not prop["select"]: return ""
return prop["select"]["name"]
def sync():
logger.info("--- Starting Enhanced Sync ---")
# 1. Init DB
init_db()
session = SessionLocal()
# 2. Sync Categories (Products)
cat_db_id = find_db_id("Product Categories") or find_db_id("Products")
if cat_db_id:
logger.info(f"Syncing Products from {cat_db_id}...")
pages = query_all(cat_db_id)
for page in pages:
props = page["properties"]
name = extract_title(props.get("Name") or props.get("Product Name"))
if not name: continue
notion_id = page["id"]
key = name.lower().replace(" ", "_")
# Upsert
cat = session.query(RoboticsCategory).filter(RoboticsCategory.notion_id == notion_id).first()
if not cat:
cat = RoboticsCategory(notion_id=notion_id, key=key)
session.add(cat)
cat.name = name
cat.description = extract_rich_text(props.get("Description"))
# Add reasoning guide map if available
session.commit()
else:
logger.warning("Product DB not found!")
# 3. Sync Industries
ind_db_id = find_db_id("Industries")
if ind_db_id:
logger.info(f"Syncing Industries from {ind_db_id}...")
# Clear existing? Or Upsert?
# For clean sync, DELETE is safer as long as we don't have FK constraints blocking it.
# But wait! Companies link to Industry STRING, not FK usually?
# Check Company model: industry_ai = Column(String). So no FK constraint. Safe to delete.
session.query(Industry).delete()
session.commit()
pages = query_all(ind_db_id)
count = 0
for page in pages:
props = page["properties"]
name = extract_title(props.get("Vertical"))
if not name: continue
ind = Industry(notion_id=page["id"], name=name)
session.add(ind)
# Map Fields
ind.description = extract_rich_text(props.get("Definition"))
ind.notes = extract_rich_text(props.get("Notes"))
ind.pains = extract_rich_text(props.get("Pains"))
ind.gains = extract_rich_text(props.get("Gains"))
# Status / Priority (Renamed field check)
# Try "Priorität" first, then "Freigegeben", then "Status"
prio = extract_select(props.get("Priorität"))
if not prio: prio = extract_select(props.get("Freigegeben"))
if not prio: prio = extract_select(props.get("Status"))
ind.priority = prio
ind.status_notion = prio # Legacy field
ind.is_focus = (prio == "Freigegeben" or prio == "P1 Focus Industry")
# Ops Focus
if "Ops Focus: Secondary" in props:
ind.ops_focus_secondary = props["Ops Focus: Secondary"].get("checkbox", False)
# Relations
# Primary
rels_prim = props.get("Primary Product Category", {}).get("relation", [])
if rels_prim:
pid = rels_prim[0]["id"]
cat = session.query(RoboticsCategory).filter(RoboticsCategory.notion_id == pid).first()
if cat: ind.primary_category_id = cat.id
# Secondary
rels_sec = props.get("Secondary Product", {}).get("relation", [])
if rels_sec:
pid = rels_sec[0]["id"]
cat = session.query(RoboticsCategory).filter(RoboticsCategory.notion_id == pid).first()
if cat: ind.secondary_category_id = cat.id
count += 1
session.commit()
logger.info(f"✅ Synced {count} industries.")
else:
logger.error("Industries DB not found!")
session.close()
if __name__ == "__main__":
sync()

102
run_lead_engine.py Normal file
View File

@@ -0,0 +1,102 @@
import argparse
import subprocess
import os
import sys
from datetime import datetime
# --- Setup Paths ---
SCRIPT_DIR = os.path.dirname(os.path.abspath(__file__))
SCRIPTS_SUBDIR = os.path.join(SCRIPT_DIR, "scripts")
LOG_DIR = os.path.join(SCRIPT_DIR, "logs")
THROUGHPUT_LOG = os.path.join(LOG_DIR, "throughput.log")
# Add scripts subdir to path to allow imports
sys.path.append(SCRIPTS_SUBDIR)
# TODO: Import other modules once they are ready
# from company_explorer_connector import handle_company_workflow
# from generate_sniper_copy import generate_copy
def setup_environment():
"""Ensures necessary directories exist."""
os.makedirs(LOG_DIR, exist_ok=True)
def log_throughput(identifier):
"""Logs a successful processing event for the dashboard."""
with open(THROUGHPUT_LOG, "a") as f:
f.write(f"{datetime.utcnow().isoformat()},{identifier}\n")
print(f"📈 Logged successful processing for '{identifier}' for dashboard.")
def run_sync():
"""Runs the database sync script to ensure local data is fresh."""
print("\n--- [Step 1: Syncing local Company Explorer database] ---")
sync_script_path = os.path.join(SCRIPTS_SUBDIR, "sync_ce_to_sqlite.py")
if not os.path.exists(sync_script_path):
print(f"❌ ERROR: Sync script not found at {sync_script_path}")
return False
result = subprocess.run(["python3", sync_script_path], capture_output=True, text=True, check=False)
if result.returncode != 0:
print("❌ ERROR: Database sync failed.")
print(result.stderr)
return False
print("✅ Sync successful.")
return True
def process_lead(identifier):
"""
Orchestrates the full enrichment and copy generation for a single lead.
"""
print(f"\n======= PROCESSING LEAD: {identifier} =======")
# --- Step 2: Enrich Company (if necessary) ---
print("\n--- [Step 2: Check/Enrich Company Data] ---")
# ce_data = handle_company_workflow(identifier) # Example of direct import
# if not ce_data or 'error' in ce_data:
# print(f"❌ Failed to enrich '{identifier}'. Aborting.")
# return
print("... (Placeholder for Enrichment Logic)")
print("✅ Enrichment complete.")
# --- Step 3: Generate Sniper Copy ---
print("\n--- [Step 3: Generate Sniper Copy] ---")
# sniper_copy = generate_copy(ce_data['data']['id'])
# print("\nGENERATED COPY:\n", sniper_copy)
print("... (Placeholder for Sniper Copy Generation)")
print("✅ Copy generation complete.")
# --- Step 4: Finalize & Log ---
print("\n--- [Step 4: Finalizing] ---")
log_throughput(identifier)
print(f"✅ Successfully processed lead '{identifier}'.")
print("============================================")
def main():
parser = argparse.ArgumentParser(description="GTM Lead Engine Orchestrator.")
parser.add_argument("leads", nargs='+', help="One or more company names or SuperOffice IDs to process.")
parser.add_argument("--skip-sync", action="store_true", help="Skip the initial database sync for faster iteration.")
args = parser.parse_args()
print("🚀 GTM Lead Engine Orchestrator started.")
setup_environment()
if not args.skip_sync:
if not run_sync():
sys.exit(1) # Exit if sync fails
else:
print("\n--- [Skipping Step 1: Database Sync] ---")
for lead_identifier in args.leads:
process_lead(lead_identifier)
print("\n🎉 Orchestration complete for all leads. 🎉")
if __name__ == "__main__":
main()

75
scripts/add_manual_report.py Normal file
View File

@@ -0,0 +1,75 @@
import os
import requests
import json
from datetime import datetime
from zoneinfo import ZoneInfo
# Configuration
NOTION_TOKEN = "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"
PAGE_ID = "2ff88f42854480008314c9013414d1d0"
BERLIN_TZ = ZoneInfo("Europe/Berlin")
def add_status_to_notion():
headers = {
"Authorization": f"Bearer {NOTION_TOKEN}",
"Content-Type": "application/json",
"Notion-Version": "2022-06-28"
}
# 1. Update the 'Total Duration (h)' field
# First, get current value
resp = requests.get(f"https://api.notion.com/v1/pages/{PAGE_ID}", headers=headers)
page_data = resp.json()
current_hours = page_data.get("properties", {}).get("Total Duration (h)", {}).get("number") or 0.0
new_hours = current_hours + 3.2
# Update property
update_payload = {
"properties": {
"Total Duration (h)": {"number": new_hours},
"Status": {"status": {"name": "Doing"}}
}
}
requests.patch(f"https://api.notion.com/v1/pages/{PAGE_ID}", headers=headers, json=update_payload)
# 2. Append the Status Update Block
timestamp = datetime.now(BERLIN_TZ).strftime('%Y-%m-%d %H:%M')
report_content = (
"Investierte Zeit in dieser Session: 03:12\n"
"Neuer Status: Doing\n\n"
"Arbeitszusammenfassung:\n"
"Wir haben heute den entscheidenden technischen Durchbruch bei der bidirektionalen Datensynchronisation zwischen dem Company Explorer (CE) und SuperOffice CRM (SO) erzielt.\n\n"
"1. Infrastruktur-Stabilisierung: Das Git-Repository wurde über eine interne Docker-Netzwerk-Verbindung (gitea-internal) stabil angebunden.\n"
"2. Pipeline-Durchstich (SO -> CE): Firmenstammdaten aus SuperOffice (Contact ID 2) werden sauber in den Company Explorer übertragen.\n"
"3. Round-Trip & Write-Back (CE -> SO): Das Protokoll für den Rückschreibeprozess wurde geknackt. Erkenntnis: SuperOffice ignoriert UrlAddress beim PUT, wir nutzen jetzt das Urls-Array. Pflichtfelder wie Number2 werden nun explizit mitgegeben."
)
block_payload = {
"children": [
{
"object": "block",
"type": "heading_2",
"heading_2": {
"rich_text": [{"type": "text", "text": {"content": f"🤖 Status-Update ({timestamp} Berlin Time)"}}]
}
},
{
"object": "block",
"type": "code",
"code": {
"rich_text": [{"type": "text", "text": {"content": report_content}}],
"language": "yaml"
}
}
]
}
final_resp = requests.patch(f"https://api.notion.com/v1/blocks/{PAGE_ID}/children", headers=headers, json=block_payload)
if final_resp.status_code == 200:
print(f"✅ SUCCESS: Notion Task updated. Total hours now: {new_hours}")
else:
print(f"❌ ERROR: {final_resp.text}")
if __name__ == "__main__":
add_status_to_notion()

115
scripts/audit_notion_consistency.py Normal file
View File

@@ -0,0 +1,115 @@
import os
import requests
import json
from dotenv import load_dotenv
load_dotenv(dotenv_path="/home/node/clawd/.env")
NOTION_TOKEN = os.getenv("NOTION_API_KEY")
HEADERS = {
"Authorization": f"Bearer {NOTION_TOKEN}",
"Content-Type": "application/json",
"Notion-Version": "2022-06-28"
}
def find_db_id(query_name):
url = "https://api.notion.com/v1/search"
payload = {"query": query_name, "filter": {"value": "database", "property": "object"}}
resp = requests.post(url, headers=HEADERS, json=payload)
if resp.status_code == 200:
results = resp.json().get("results", [])
if results:
return results[0]['id']
return None
# Cache for product names to avoid API spam
product_cache = {}
def resolve_product_name(relation_ids):
if not relation_ids:
return "None"
names = []
for rel in relation_ids:
page_id = rel['id']
if page_id in product_cache:
names.append(product_cache[page_id])
continue
url = f"https://api.notion.com/v1/pages/{page_id}"
resp = requests.get(url, headers=HEADERS)
if resp.status_code == 200:
props = resp.json().get("properties", {})
# Assume Product DB has a Title field called "Name" or "Product Name"
# We iterate to find the title
title = "Unknown"
for key, val in props.items():
if val['id'] == 'title': # The title property always has id 'title'
if val['title']:
title = val['title'][0]['plain_text']
break
product_cache[page_id] = title
names.append(title)
else:
names.append("Error fetching Product")
return ", ".join(names)
def audit_industries():
db_id = find_db_id("Industries")
if not db_id:
print("❌ Industries DB not found.")
return
print(f"--- Auditing Industries DB ({db_id}) ---")
url = f"https://api.notion.com/v1/databases/{db_id}/query"
resp = requests.post(url, headers=HEADERS, json={})
if resp.status_code != 200:
print(f"Error: {resp.text}")
return
pages = resp.json().get("results", [])
# We want to see: Vertical Name | Status | Primary Product (Resolved) | Notes Snippet
print(f"{'Vertical':<35} | {'Status':<15} | {'Primary Product':<30} | {'Notes (Snippet)'}")
print("-" * 120)
for page in pages:
props = page['properties']
# Name
name = "N/A"
if "Vertical" in props and props["Vertical"]["title"]:
name = props["Vertical"]["title"][0]["plain_text"]
elif "Name" in props and props["Name"]["title"]: # Fallback
name = props["Name"]["title"][0]["plain_text"]
# Filter for the ones we touched or are interested in
# (Optional: remove filter to see all)
# Status
status = ""
if "Freigabe" in props:
if props["Freigabe"]["type"] == "status" and props["Freigabe"]["status"]:
status = props["Freigabe"]["status"]["name"]
elif props["Freigabe"]["type"] == "select" and props["Freigabe"]["select"]:
status = props["Freigabe"]["select"]["name"]
# Primary Product (Relation)
product_name = "None"
if "Primary Product Category" in props and props["Primary Product Category"]["relation"]:
product_name = resolve_product_name(props["Primary Product Category"]["relation"])
# Notes
notes = ""
if "Notes" in props and props["Notes"]["rich_text"]:
full_note = props["Notes"]["rich_text"][0]["plain_text"]
notes = (full_note[:40] + '...') if len(full_note) > 40 else full_note
if name != "N/A":
print(f"{name:<35} | {status:<15} | {product_name:<30} | {notes}")
if __name__ == "__main__":
audit_industries()

27
scripts/discover_notion_dbs.py Normal file
View File

@@ -0,0 +1,27 @@
import os
import requests
import json
from dotenv import load_dotenv
load_dotenv(dotenv_path="/home/node/clawd/.env")
def discover_dbs():
token = os.getenv("NOTION_API_KEY")
headers = {
"Authorization": f"Bearer {token}",
"Content-Type": "application/json",
"Notion-Version": "2022-06-28"
}
url = "https://api.notion.com/v1/search"
payload = {"filter": {"value": "database", "property": "object"}}
resp = requests.post(url, headers=headers, json=payload)
results = resp.json().get("results", [])
print("--- Gefundene Datenbanken ---")
for db in results:
title = db.get("title", [{}])[0].get("plain_text", "Unbekannt")
print(f"Name: {title} | ID: {db['id']}")
if __name__ == "__main__":
discover_dbs()

265
scripts/enrich_notion_pains.py Normal file
View File

@@ -0,0 +1,265 @@
import os
import requests
import json
from dotenv import load_dotenv
load_dotenv(dotenv_path="/home/node/clawd/.env")
NOTION_TOKEN = os.getenv("NOTION_API_KEY")
HEADERS = {
"Authorization": f"Bearer {NOTION_TOKEN}",
"Content-Type": "application/json",
"Notion-Version": "2022-06-28"
}
# --- Load Product Mapping ---
try:
with open("data/product_mapping.json", "r") as f:
PRODUCT_MAP = json.load(f)
except FileNotFoundError:
print("❌ Product mapping not found. Run fetch_product_mapping.py first.")
exit(1)
# Helper to find DB ID
def find_db_id(query_name):
url = "https://api.notion.com/v1/search"
payload = {"query": query_name, "filter": {"value": "database", "property": "object"}}
resp = requests.post(url, headers=HEADERS, json=payload)
if resp.status_code == 200:
results = resp.json().get("results", [])
if results:
return results[0]['id']
return None
def get_page_by_vertical(db_id, vertical_name):
url = f"https://api.notion.com/v1/databases/{db_id}/query"
# Using 'Vertical' as the title property name based on previous audit
payload = {
"filter": {
"property": "Vertical",
"title": {"equals": vertical_name}
}
}
resp = requests.post(url, headers=HEADERS, json=payload)
if resp.status_code == 200:
results = resp.json().get("results", [])
if results:
return results[0]
return None
def update_page(page_id, properties):
url = f"https://api.notion.com/v1/pages/{page_id}"
payload = {"properties": properties}
resp = requests.patch(url, headers=HEADERS, json=payload)
if resp.status_code == 200:
print(f"✅ Updated '{page_id}'")
else:
print(f"❌ Error updating '{page_id}': {resp.text}")
def create_page(db_id, properties):
url = "https://api.notion.com/v1/pages"
payload = {"parent": {"database_id": db_id}, "properties": properties}
resp = requests.post(url, headers=HEADERS, json=payload)
if resp.status_code == 200:
print(f"✅ Created new page")
else:
print(f"❌ Error creating page: {resp.text}")
# --- CONTENT DEFINITION ---
# Format: Vertical -> { props... }
UPDATES = {
"Healthcare - Care Home": {
"product": "Cleaning Indoor Roboter (Wet Surface)",
"pains": """[Primary Product: Cleaning]
- Infektionsrisiko: Mangelnde Bodenhygiene und Keimverschleppung in sensiblen Bereichen gefährden Bewohner.
- Dokumentationspflicht: Lückenlose Nachweise für Hygiene-Audits binden wertvolle Pflegezeit.
- Personalmangel: Reinigungskräfte fehlen, Standards können manuell kaum gehalten werden.
[Secondary Product: Service]
- Pflegeressourcen: Fachkräfte binden bis zu 30% ihrer Zeit mit nicht-pflegerischen Transportwegen (Essen/Wäsche).
- Körperliche Belastung: Schweres Heben und weite Wege führen zu krankheitsbedingten Ausfällen im Pflegeteam.""",
"gains": """[Primary Product: Cleaning]
- Audit-Sicherheit: Automatisierte, protokollierte Reinigung sichert Compliance ohne Mehraufwand.
- Entlastung Housekeeping: Personal konzentriert sich auf Sichtreinigung und Desinfektion statt Bodenfläche.
[Secondary Product: Service]
- Mehr Zeit am Patienten: Reduktion der Laufwege gibt Pflegekräften 2-3 Std./Schicht zurück.
- Mitarbeiterzufriedenheit: Reduktion körperlicher Belastung senkt Krankenstand.""",
"ops_focus": True,
"status": "Freigegeben",
"notes": "Prio 1: Reinigung. Prio 2: Service (Essen). Fokus auf Fachkräftemangel & Hygiene."
},
"Healthcare - Hospital": {
"product": "Cleaning Indoor Roboter (Wet Surface)",
"pains": """[Primary Product: Cleaning]
- Infektionsschutz: Hohe Frequenz an Patientenbewegungen erfordert permanente Desinfektion der Böden.
- Audit-Druck: Behördliche Auflagen verlangen lückenlose Dokumentation, die manuell kaum leistbar ist.
- Kostendruck: Steigende Personalkosten bei fixen Fallpauschalen zwingen zur Effizienzsteigerung.
[Secondary Product: Service]
- Logistik-Aufwand: Transport von Proben, Wäsche und Essen bindet Pflegepersonal in unproduktiven Wegezeiten.""",
"gains": """[Primary Product: Cleaning]
- Hygiene-Standard: 24/7 gleichbleibende Reinigungsqualität reduziert Keimbelastung messbar.
- Compliance: Automatische Protokollierung aller Reinigungsfahrten für Audits.
[Secondary Product: Service]
- Prozess-Effizienz: Automatisierter Warentransport entlastet Fachpersonal für medizinische Aufgaben.""",
"ops_focus": True,
"status": "Freigegeben",
"notes": "Prio 1: Reinigung (Alex Veto). Service ist 'nice to have'. KPI: Hygiene-Sicherheit."
},
"Leisure - Entertainment": {
"product": "Service Roboter", # FIX: Changed from Cleaning to Service
"pains": """[Primary Product: Service]
- Service-Engpass: Umsatzverlust zu Stoßzeiten, da Personal nicht schnell genug Getränke/Snacks nachliefert.
- Personalmangel: Schwierige Besetzung von Spätschichten führt zu geschlossenen Stationen/Bahnen.
- Wartezeiten: Gäste sind unzufrieden, wenn Bestellungen zu lange dauern.
[Secondary Product: Cleaning]
- Bodenverschmutzung: Klebrige Böden (Getränke/Popcorn) im Foyer stören das Gästeerlebnis.""",
"gains": """[Primary Product: Service]
- Umsatzsteigerung: Permanente Verfügbarkeit von Snacks/Getränken direkt am Platz (Cross-Selling).
- Erlebnis-Faktor: Innovative Roboter begeistern Gäste und fördern Social-Media-Sichtbarkeit.
- Entlastung: Servicepersonal hat mehr Zeit für Gästebetreuung statt Laufwege.""",
"ops_focus": True, # Keep secondary focus plausible
"status": "Freigegeben",
"notes": "Prio 1: Service Robotik (BellaBot). Cleaning nur Prio 2 (Foyer/Gänge)."
},
"Logistics - Warehouse": {
"product": "Cleaning Outdoor Roboter (Sweeper)",
"pains": """[Primary Product: Sweeper]
- Prozesssicherheit: Staub auf Sensoren und Lichtschranken führt zu Anlagenstörungen und Produktionsstopps.
- Arbeitssicherheit: Verschmutzte Fahrwege durch Palettenreste/Staub erhöhen das Unfallrisiko.
- Manuelle Bindung: Fachkräfte müssen kehren statt kommissionieren.
[Secondary Product: Cleaning Wet]
- Hartnäckige Verschmutzungen: Öl/Reifenabrieb erfordern Nassreinigung, die manuell zeitintensiv ist.""",
"gains": """[Primary Product: Sweeper]
- Staubfreie Umgebung: Werterhalt des Hallenbodens und Schutz empfindlicher Ware/Anlagen.
- Produktivität: Reinigung erfolgt parallel zum Betrieb oder nachts, ohne Störung.
- Sicherheit: Saubere Fahrwege reduzieren Unfallrisiko für Flurförderzeuge.""",
"ops_focus": True,
"status": "Freigegeben",
"notes": "Prio 1: Sweeper (Staub). Prio 2: Wet. Transport schwierig wegen Paletten."
},
"Tech - Data Center": {
"product": "Security Roboter",
"pains": """[Primary Product: Security]
- Sicherheitsrisiko: Unbefugter Zutritt in Sicherheitsbereiche muss lückenlos detektiert werden (24/7).
- Personalbindung: Wachpersonal ist teuer und kann nicht überall gleichzeitig sein.
[Secondary Product: Cleaning]
- Feinstaub: Staubpartikel in Serverräumen gefährden Hardware und Kühlung.""",
"gains": """[Primary Product: Security]
- Lückenlose Überwachung: Permanente Patrouille und sofortige Alarmierung ohne Personalbindung.
- Dokumentation: Video- und Sensorprotokolle aller Ereignisse.
[Secondary Product: Cleaning]
- Ausfallsicherheit: Staubfreie Umgebung verlängert Hardware-Lebensdauer.""",
"ops_focus": True,
"status": "Klärrungsbedarf", # New, needs review
"notes": "Neu angelegt. Prio 1 Security (lt. Transkript). Prio 2 Cleaning (Staub)."
},
"Reinigungsdienstleister": {
"product": "Cleaning Indoor Roboter (Wet Surface)",
"pains": """[Primary Product: Cleaning]
- Personalmangel: Schwierigkeit, zuverlässiges Personal für alle Objekte zu finden.
- Kostendruck: Geringe Margen bei Ausschreibungen erfordern hohe Effizienz.
- Qualitätsschwankungen: Manuelle Reinigung variiert stark, Kunden beschweren sich.
- Fluktuation: Hoher Aufwand für ständige Neueinarbeitung.""",
"gains": """[Primary Product: Cleaning]
- Skalierbarkeit: Roboter übernehmen Flächenleistung, Personal macht Detailreinigung.
- Innovation: Wettbewerbsvorteil bei Ausschreibungen durch Technologie-Einsatz.
- Kalkulationssicherheit: Fixe Kosten statt variabler Personalkosten/Krankheitstage.""",
"ops_focus": False,
"status": "Klärrungsbedarf",
"notes": "Neu angelegt. Zielgruppe: Wisag, Dussmann etc. (Alex: Größter Markt)."
},
"Infrastructure - Communities": {
"product": "Cleaning Indoor Roboter (Wet Surface)",
"pains": """[Primary Product: Cleaning]
- Großflächen-Reinigung: Sporthallen, Aulen und Flure binden enorm viele Personalstunden.
- Budget-Druck: Kommunen müssen sparen, Reinigungskosten sind großer Posten.
- Nutzungs-Konflikte: Reinigung muss in engen Zeitfenstern zwischen Schul/Vereinsnutzung erfolgen.""",
"gains": """[Primary Product: Cleaning]
- Kosteneffizienz: Reduktion der Reinigungskosten pro Quadratmeter.
- Flexibilität: Reinigung kann nachts oder in Randzeiten erfolgen.
- Werterhalt: Schonende, regelmäßige Reinigung verlängert Lebensdauer von Sportböden.""",
"ops_focus": False,
"status": "Klärrungsbedarf",
"notes": "Neu angelegt (Schulen, Gemeinden)."
},
"Infrastructure Parking": {
"product": "Cleaning Outdoor Roboter (Sweeper)",
"pains": """[Primary Product: Sweeper]
- Außenwirkung: Verschmutzte Parkflächen/Müll schaden dem Image (erster Eindruck).
- Manuelle Arbeit: Fegen von großen Außenflächen ist personalintensiv und unbeliebt.
- Umwelt: Müll gelangt in die Umgebung/Kanalisation.""",
"gains": """[Primary Product: Sweeper]
- Gepflegtes Erscheinungsbild: Täglich saubere Außenanlagen.
- Autonomie: Roboter reinigt selbstständig, auch bei schlechtem Wetter.
- Entlastung: Hausmeister kann sich um Wartung kümmern statt Fegen.""",
"ops_focus": False,
"status": "Klärrungsbedarf",
"notes": "Neu angelegt (Parkplätze, Außenanlagen)."
}
}
def run_enrichment():
db_id = find_db_id("Industries")
if not db_id:
print("❌ Industries DB not found.")
return
print(f"--- Enriching Verticals in DB {db_id} ---")
for vertical, data in UPDATES.items():
# Resolve Product ID
prod_id = PRODUCT_MAP.get(data["product"])
if not prod_id:
print(f"❌ Product '{data['product']}' not found in map. Skipping {vertical}.")
continue
# Prepare Properties
props = {
"Pains": {"rich_text": [{"text": {"content": data["pains"]}}]},
"Gains": {"rich_text": [{"text": {"content": data["gains"]}}]},
"Primary Product Category": {"relation": [{"id": prod_id}]},
"Notes": {"rich_text": [{"text": {"content": data["notes"]}}]},
# Handle Status (Try Select first, then Status)
# We assume "Freigabe" exists
}
# Add checkbox if present in logic
if "ops_focus" in data:
props["Ops Focus: Secondary"] = {"checkbox": data["ops_focus"]}
# Check if page exists
page = get_page_by_vertical(db_id, vertical)
if page:
# Update existing
# Add Status Update
# (Note: Logic to detect Select vs Status type is needed, but we assume Select/Status name is consistent)
# For robustness, we check the property type in the page object
status_type = page['properties'].get("Freigabe", {}).get("type")
if status_type == "status":
props["Freigabe"] = {"status": {"name": data["status"]}}
elif status_type == "select":
props["Freigabe"] = {"select": {"name": data["status"]}}
print(f"Updating '{vertical}'...")
update_page(page['id'], props)
else:
# Create new
print(f"Creating new vertical '{vertical}'...")
props["Vertical"] = {"title": [{"text": {"content": vertical}}]}
# Guess status type (usually Select or Status) - Try Status first as default in new Notion DBs
# Or omit status if unsure, but we want to set it.
# We'll try Status format.
props["Freigabe"] = {"status": {"name": data["status"]}}
create_page(db_id, props)
if __name__ == "__main__":
run_enrichment()

65
scripts/fetch_product_mapping.py Normal file
View File

@@ -0,0 +1,65 @@
import os
import requests
import json
from dotenv import load_dotenv
load_dotenv(dotenv_path="/home/node/clawd/.env")
NOTION_TOKEN = os.getenv("NOTION_API_KEY")
HEADERS = {
"Authorization": f"Bearer {NOTION_TOKEN}",
"Content-Type": "application/json",
"Notion-Version": "2022-06-28"
}
def find_db_id(query_name):
url = "https://api.notion.com/v1/search"
payload = {"query": query_name, "filter": {"value": "database", "property": "object"}}
resp = requests.post(url, headers=HEADERS, json=payload)
if resp.status_code == 200:
results = resp.json().get("results", [])
if results:
return results[0]['id']
return None
def fetch_products():
# Find Product DB (it's likely named "Product Categories" or similar based on schema)
# Or search for "Products"
db_id = find_db_id("Product Categories")
if not db_id:
db_id = find_db_id("Products") # Fallback
if not db_id:
print("❌ Could not find Product Database.")
return
print(f"--- Fetching Products from DB {db_id} ---")
url = f"https://api.notion.com/v1/databases/{db_id}/query"
resp = requests.post(url, headers=HEADERS, json={})
products = {}
if resp.status_code == 200:
results = resp.json().get("results", [])
for page in results:
props = page['properties']
# Find Title
name = "Unknown"
if "Name" in props and props["Name"]["title"]:
name = props["Name"]["title"][0]["plain_text"]
elif "Product Name" in props and props["Product Name"]["title"]:
name = props["Product Name"]["title"][0]["plain_text"]
products[name] = page['id']
print(f"Found: {name} ({page['id']})")
# Save to file
os.makedirs("data", exist_ok=True)
with open("data/product_mapping.json", "w") as f:
json.dump(products, f, indent=4)
print("✅ Saved to data/product_mapping.json")
else:
print(f"Error: {resp.text}")
if __name__ == "__main__":
fetch_products()

93
scripts/generate_ce_hooks.py Normal file
View File

@@ -0,0 +1,93 @@
import sqlite3
import os
import json
import requests
from dotenv import load_dotenv
# Load ENV for Gemini API
load_dotenv(dotenv_path="/home/node/clawd/.env", override=True)
class LeadHookService:
def __init__(self, db_path):
self.db_path = db_path
self.api_key = os.getenv("GEMINI_API_KEY")
def _get_company_data(self, company_id):
conn = sqlite3.connect(self.db_path)
conn.row_factory = sqlite3.Row
cursor = conn.cursor()
# Get company and metrics
cursor.execute("SELECT * FROM companies WHERE id = ?", (company_id,))
company = cursor.fetchone()
if not company:
return None
data = dict(company)
conn.close()
return data
def build_combined_context(self, company_data):
# Build the 'combined' string from CE facts
parts = []
parts.append(f"Name: {company_data.get('name')}")
parts.append(f"Branche: {company_data.get('industry_ai')}")
if company_data.get('calculated_metric_value'):
parts.append(f"Metrik: {company_data.get('calculated_metric_value')} {company_data.get('calculated_metric_unit')}")
# Add a hint about the core business from status/city
parts.append(f"Standort: {company_data.get('city')}")
return " | ".join(parts)
def generate_hook(self, company_id):
company_data = self._get_company_data(company_id)
if not company_data:
return "Company not found."
combined = self.build_combined_context(company_data)
display_name = company_data.get('name').split(' ')[0] # Simple Kurzname logic
prompt = f"""
Du bist ein exzellenter B2B-Stratege und Texter.
Deine Aufgabe ist es, einen hochpersonalisierten, scharfsinnigen und wertschätzenden Einleitungssatz für eine E-Mail zu formulieren.
--- Unternehmenskontext ---
Kurzname: {display_name}
Beschreibung: {combined}
--- Stilvorgaben ---
1. Analysiere das Kerngeschäft: Was ist die zentrale physische Herausforderung (z.B. Sauberkeit in Nassbereichen, Logistik-Effizienz, Objektschutz)?
2. KEINE ZAHLEN: Erwähne niemals konkrete Zahlen (Besucherzahlen, m², Anzahl Pools). Nutze stattdessen qualitative Begriffe wie "weitläufig", "hochfrequent", "komplex" oder "marktführend".
3. Identifiziere den Hebel: Was ist der Erfolgsfaktor (z.B. Gäste-Zufriedenheit, Prozessstabilität, Sicherheit)?
4. Formuliere den Satz (20-35 Wörter): Elegant, aktiv, KEINE Floskeln.
5. WICHTIG: Formuliere als positive Beobachtung über eine Kernkompetenz.
Deine Ausgabe: NUR der finale Satz.
"""
# Call Gemini (Simplified for POC)
headers = {"Content-Type": "application/json"}
payload = {
"contents": [{"parts": [{"text": prompt}]}]
}
url = f"https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash:generateContent?key={self.api_key}"
resp = requests.post(url, headers=headers, json=payload)
result = resp.json()
try:
hook_text = result['candidates'][0]['content']['parts'][0]['text'].strip()
return hook_text
except:
return f"Error generating hook: {result}"
if __name__ == "__main__":
# Test with CE-ID 1 (Therme Erding)
db = "/home/node/clawd/repos/brancheneinstufung2/company_explorer_local.db"
service = LeadHookService(db)
print(f"--- Testing LeadHookService for ID 1 ---")
hook = service.generate_hook(1)
print(f"GENERATED HOOK:\n{hook}")

123
scripts/generate_sniper_copy.py Normal file
View File

@@ -0,0 +1,123 @@
import sqlite3
import os
import json
import requests
import argparse
from dotenv import load_dotenv
# --- Configuration & Setup ---
load_dotenv(dotenv_path="/home/node/clawd/.env", override=True)
DB_PATH = "/home/node/clawd/repos/brancheneinstufung2/company_explorer_local.db"
GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
class SniperGenerator:
def __init__(self, db_path=DB_PATH):
self.db_path = db_path
# TODO: Initialize Notion client to get Vertical/Persona data
# TODO: Load Marketing Report KPIs into memory
def get_lead_data(self, company_id):
"""Gathers all necessary data for a lead from the local DB."""
conn = sqlite3.connect(self.db_path)
conn.row_factory = sqlite3.Row
cursor = conn.cursor()
cursor.execute("SELECT * FROM companies WHERE id = ?", (company_id,))
company_data = cursor.fetchone()
conn.close()
if not company_data:
return None
return dict(company_data)
def select_product_and_persona(self, company_data, target_role):
"""
Implements the '3+1' rule to decide which product to pitch.
Placeholder logic - will be replaced with Notion data.
"""
print(f"🎯 Selecting product for role '{target_role}' in industry '{company_data.get('industry_ai')}'...")
# Placeholder for the 3+1 logic
# if target_role in ["Wirtschaftl. Entscheider", "Infrastruktur-Verantw."]:
# return "Primary"
# if target_role == "Innovations-Treiber":
# return "Secondary"
# if target_role == "Operativer Entscheider":
# # Here we would check the "Ops Focus: Secondary?" checkbox from Notion
# return "Primary" # Default
# For now, we default to the primary product (Cleaning)
print("-> Defaulting to 'Primary Product' (Cleaning).")
return "Cleaning"
def generate_copy(self, company_id, target_role="Wirtschaftl. Entscheider"):
"""
Generates the 3-sentence sniper copy for a given company and role.
"""
# 1. Gather Data
lead_data = self.get_lead_data(company_id)
if not lead_data:
return "Error: Company data not found."
# 2. Decide on Product (using 3+1 rule)
product_to_pitch = self.select_product_and_persona(lead_data, target_role)
# 3. Get Social Proof KPIs (from Marketing Report)
# Placeholder - using hardcoded values from the report
kpis = {
"cost_reduction": "10-25%",
"time_saving": "20-40%"
}
# 4. Construct Master Prompt
# This is a simplified version for now
prompt = f"""
Du bist ein Weltklasse B2B-Stratege. Deine Aufgabe ist es, eine 3-Satz-E-Mail-Einleitung im '1,5°-Stil' zu erstellen.
**Regeln:**
- Satz 1 (Firma): Zeige, dass du das Geschäftsmodell und die zentrale Herausforderung verstanden hast. KEINE ZAHLEN, nur qualitative Größe.
- Satz 2 (Persona): Sprich den spezifischen Schmerz der Zielrolle an und verbinde ihn mit dem Produkt '{product_to_pitch}'.
- Satz 3 (Social Proof): Untermauere die Lösung mit einem konkreten KPI von Marktbegleitern.
**Daten:**
- Firma: {lead_data.get('name')}
- Branche (KI): {lead_data.get('industry_ai')}
- Standort: {lead_data.get('city')}
- Rolle: {target_role}
- KPI 1 (Kosten): {kpis['cost_reduction']}
- KPI 2 (Zeit): {kpis['time_saving']}
**Output:** Nur die 3 Sätze. Sonst nichts.
"""
# 5. Call Gemini API
print("📞 Calling Gemini to generate copy...")
headers = {"Content-Type": "application/json"}
payload = {"contents": [{"parts": [{"text": prompt}]}]}
url = f"https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash:generateContent?key={GEMINI_API_KEY}"
try:
resp = requests.post(url, headers=headers, json=payload, timeout=20)
resp.raise_for_status()
result = resp.json()
copy_text = result['candidates'][0]['content']['parts'][0]['text'].strip()
return copy_text
except Exception as e:
return f"Error during Gemini call: {e}"
if __name__ == "__main__":
parser = argparse.ArgumentParser(description="Generate sniper copy for a lead.")
parser.add_argument("company_id", type=int, help="The Company Explorer ID of the lead.")
parser.add_argument("--role", type=str, default="Wirtschaftl. Entscheider", help="The target persona/role.")
args = parser.parse_args()
sniper = SniperGenerator()
final_copy = sniper.generate_copy(args.company_id, args.role)
print("\n--- GENERATED SNIPER COPY ---")
print(final_copy)
print("-----------------------------\n")

106
scripts/query_notion_db.py Normal file
View File

@@ -0,0 +1,106 @@
import os
import requests
import json
from dotenv import load_dotenv
load_dotenv(dotenv_path="/home/node/clawd/.env")
def find_db_by_name(query_name):
token = os.getenv("NOTION_API_KEY")
headers = {
"Authorization": f"Bearer {token}",
"Content-Type": "application/json",
"Notion-Version": "2022-06-28"
}
url = "https://api.notion.com/v1/search"
payload = {
"query": query_name,
"filter": {"value": "database", "property": "object"}
}
# print(f"Searching for '{query_name}' database...")
resp = requests.post(url, headers=headers, json=payload)
if resp.status_code != 200:
print(f"Error searching DB: {resp.text}")
return None
results = resp.json().get("results", [])
if not results:
# print(f"No database named '{query_name}' found via search.")
return None
db = results[0]
return db['id']
def dump_db_content(db_id, db_name="DB"):
token = os.getenv("NOTION_API_KEY")
headers = {
"Authorization": f"Bearer {token}",
"Content-Type": "application/json",
"Notion-Version": "2022-06-28"
}
# Get all pages
url = f"https://api.notion.com/v1/databases/{db_id}/query"
resp = requests.post(url, headers=headers, json={})
if resp.status_code != 200:
print(f"Error querying DB: {resp.text}")
return
pages = resp.json().get("results", [])
print(f"\n--- Content of '{db_name}' ({len(pages)} rows) ---")
rows = []
for page in pages:
props = page['properties']
# Extract Name (Title) - Robust Logic
name = "N/A"
if "Vertical" in props and props["Vertical"]["title"]:
name = props["Vertical"]["title"][0]["plain_text"]
elif "Name" in props and props["Name"]["title"]:
name = props["Name"]["title"][0]["plain_text"]
elif "Role" in props and props["Role"]["title"]:
name = props["Role"]["title"][0]["plain_text"]
# Extract Status/Freigabe
freigabe = ""
if "Freigabe" in props:
if props["Freigabe"]["type"] == "status":
freigabe = props["Freigabe"]["status"]["name"] if props["Freigabe"]["status"] else ""
elif props["Freigabe"]["type"] == "select":
freigabe = props["Freigabe"]["select"]["name"] if props["Freigabe"]["select"] else ""
# Extract Notes
notes = ""
if "Notes" in props and props["Notes"]["rich_text"]:
notes = props["Notes"]["rich_text"][0]["plain_text"]
# Extract KPIs
kpis = ""
for kpi_key in ["KPIs", "KPI", "Quantitative Value"]:
if kpi_key in props and props[kpi_key]["rich_text"]:
kpis = props[kpi_key]["rich_text"][0]["plain_text"]
break
rows.append({"name": name, "freigabe": freigabe, "notes": notes, "kpis": kpis})
# Print clean table
print(f"{'Name':<40} | {'Freigabe':<15} | {'KPIs':<20} | {'Notes'}")
print("-" * 120)
for r in rows:
# Nur Zeilen mit Inhalt anzeigen (Filter empty names)
if r['name'] != "N/A":
print(f"{r['name']:<40} | {r['freigabe']:<15} | {r['kpis']:<20} | {r['notes']}")
if __name__ == "__main__":
db_id_ind = find_db_by_name("Industries")
if db_id_ind:
dump_db_content(db_id_ind, "Industries")
db_id_roles = find_db_by_name("Personas")
if db_id_roles:
dump_db_content(db_id_roles, "Personas")

130
scripts/sync_ce_to_sqlite.py Normal file
View File

@@ -0,0 +1,130 @@
import requests
import os
import sqlite3
import json
# --- Configuration ---
BASE_URL = "http://192.168.178.6:8090/ce/api"
API_USER = os.getenv("COMPANY_EXPLORER_API_USER", "admin")
API_PASSWORD = os.getenv("COMPANY_EXPLORER_API_PASSWORD", "gemini")
DB_PATH = "/home/node/clawd/repos/brancheneinstufung2/company_explorer_local.db"
def fetch_all_companies_from_api():
"""Fetches all companies from the Company Explorer API."""
print("Fetching all companies from Company Explorer API...")
url = f"{BASE_URL}/companies"
all_companies = []
page = 1
while True:
try:
params = {'page': page, 'per_page': 50}
response = requests.get(
url,
auth=(API_USER, API_PASSWORD),
params=params,
timeout=20
)
response.raise_for_status()
data = response.json()
companies_on_page = data.get("items", [])
if not companies_on_page:
break
all_companies.extend(companies_on_page)
print(f"Fetched page {page} with {len(companies_on_page)} companies.")
if len(all_companies) >= data.get("total", 0):
break
page += 1
except requests.exceptions.RequestException as e:
print(f"Error fetching companies from API: {e}")
return None
print(f"Total companies fetched: {len(all_companies)}")
return all_companies
def setup_database():
"""Creates the SQLite database and the companies table."""
print(f"Setting up database at: {DB_PATH}")
if os.path.exists(DB_PATH):
os.remove(DB_PATH)
print("Removed existing database file.")
conn = sqlite3.connect(DB_PATH)
cursor = conn.cursor()
# Define a flexible schema to hold the key fields
cursor.execute("""
CREATE TABLE companies (
id INTEGER PRIMARY KEY,
name TEXT,
industry_ai TEXT,
status TEXT,
city TEXT,
country TEXT,
website TEXT,
calculated_metric_name TEXT,
calculated_metric_value TEXT,
calculated_metric_unit TEXT,
full_json TEXT
)
""")
conn.commit()
conn.close()
print("Database and table 'companies' created successfully.")
def populate_database(companies):
"""Populates the database with company data."""
if not companies:
print("No companies to populate.")
return
print("Populating database...")
conn = sqlite3.connect(DB_PATH)
cursor = conn.cursor()
for company in companies:
# Extract metrics safely
metrics = company.get('calculated_metrics', [])
metric_name = metrics[0].get('name') if metrics else None
metric_value = metrics[0].get('value') if metrics else None
metric_unit = metrics[0].get('unit') if metrics else None
cursor.execute("""
INSERT INTO companies (
id, name, industry_ai, status, city, country, website,
calculated_metric_name, calculated_metric_value, calculated_metric_unit,
full_json
) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
""", (
company.get('id'),
company.get('name'),
company.get('industry_ai'),
company.get('status'),
company.get('city'),
company.get('country'),
company.get('website'),
metric_name,
metric_value,
metric_unit,
json.dumps(company) # Store the full object for future flexibility
))
conn.commit()
conn.close()
print(f"Successfully inserted {len(companies)} records into the database.")
if __name__ == "__main__":
all_companies = fetch_all_companies_from_api()
if all_companies is not None:
setup_database()
populate_database(all_companies)
print("\nSync process finished successfully.")
print(f"Database is ready at: {DB_PATH}")
else:
print("\nSync process failed due to API errors.")

25
scripts/test_prod_key.py Normal file
View File

@@ -0,0 +1,25 @@
import os
import requests
from dotenv import load_dotenv
load_dotenv(dotenv_path="/home/node/clawd/.env")
def test_prod_key():
key = os.getenv("GEMINI_API_KEY_PROD")
if not key:
print("❌ GEMINI_API_KEY_PROD not found in .env")
return
print(f"🔑 Testing Key: {key[:5]}...{key[-3:]}")
url = f"https://generativelanguage.googleapis.com/v1beta/models?key={key}"
resp = requests.get(url)
if resp.status_code == 200:
print("✅ API Call Successful! Key is active.")
# print(f"Available Models: {[m['name'] for m in resp.json().get('models', [])][:3]}")
else:
print(f"❌ API Error: {resp.status_code} - {resp.text}")
if __name__ == "__main__":
test_prod_key()

200
scripts/update_notion_batch.py Normal file
View File

@@ -0,0 +1,200 @@
import os
import requests
import json
from dotenv import load_dotenv
load_dotenv(dotenv_path="/home/node/clawd/.env")
NOTION_TOKEN = os.getenv("NOTION_API_KEY")
HEADERS = {
"Authorization": f"Bearer {NOTION_TOKEN}",
"Content-Type": "application/json",
"Notion-Version": "2022-06-28"
}
def find_db_id(query_name):
url = "https://api.notion.com/v1/search"
payload = {"query": query_name, "filter": {"value": "database", "property": "object"}}
resp = requests.post(url, headers=HEADERS, json=payload)
if resp.status_code == 200:
results = resp.json().get("results", [])
if results:
return results[0]['id']
return None
def get_page_id(db_id, title_col, title_val):
url = f"https://api.notion.com/v1/databases/{db_id}/query"
payload = {
"filter": {
"property": title_col,
"title": {"equals": title_val}
}
}
resp = requests.post(url, headers=HEADERS, json=payload)
if resp.status_code == 200:
results = resp.json().get("results", [])
if results:
return results[0] # Return full page obj to access props
return None
def update_page(page_id, properties):
url = f"https://api.notion.com/v1/pages/{page_id}"
payload = {"properties": properties}
resp = requests.patch(url, headers=HEADERS, json=payload)
if resp.status_code == 200:
print(f"✅ Updated page {page_id}")
else:
print(f"❌ Error updating {page_id}: {resp.text}")
def append_text(current_text, new_text):
if not current_text:
return new_text
if new_text in current_text:
return current_text # Avoid duplicates
return f"{current_text}\n\n[Auto-Update]: {new_text}"
# --- DATA TO UPDATE ---
# 1. Personas (KPIs)
PERSONA_UPDATES = {
"Wirtschaftlicher Entscheider": "10-25% Reduktion Personalkosten\n15-30% höhere Gästezufriedenheit (Hypothese)",
"Operativer Entscheider": "20-40% Entlastung bei Routineaufgaben\n100% Abdeckung Reinigungszyklen",
"Infrastruktur-Verantwortlicher": "20-30% schnellere Integration\n80-90% weniger Ausfallzeiten",
"Innovations-Treiber": "10-20% höhere Servicekapazität\nSteigerung Upselling 5-10%"
}
# 2. Industries (Pains/Gains/Status/Notes)
INDUSTRY_UPDATES = {
"Healthcare - Hospital": {
"pains_add": "Mangelnde Hygiene-Standards durch Personalengpässe (Infektionsrisiko). Hoher Dokumentationsaufwand für Audits.",
"gains_add": "Konstante, audit-sichere Sauberkeit (24/7). Entlastung des Reinigungspersonals.",
"status": "Freigegeben",
"note_add": "Prio 1: Reinigung (Alex Veto). Service ist 'nice to have'. KPI: Hygiene-Sicherheit.",
"ops_focus": True # Checkbox
},
"Healthcare - Care Home": {
"pains_add": "Mangelnde Hygiene-Standards. Steigende Personalkosten bei begrenzten Pflegesätzen.",
"gains_add": "Sichtbare Hygiene schafft Vertrauen. Entlastung Housekeeping.",
"status": "Freigegeben",
"note_add": "Prio 1: Reinigung. Prio 2: Service (Essen). Fokus auf Fachkräftemangel.",
"ops_focus": True
},
"Hospitality - Gastronomy": {
"pains_add": "Lobby-Optik leidet bei Personalmangel.",
"gains_add": "Makellose Optik für den ersten Eindruck.",
"status": "Freigegeben",
"note_add": "Prio 1: Reinigung (Nachts). Service nur in Entertainment-Gastro.",
"ops_focus": False
},
"Leisure - Entertainment": {
"pains_add": "Service-Personal fehlt für Umsatz (Getränke).",
"gains_add": "Mehr Umsatz durch konstante Verfügbarkeit.",
"status": "Freigegeben",
"note_add": "Prio 1: Service Robotik (BellaBot).",
"ops_focus": False
},
"Industry - Manufacturing": {
"pains_add": "Staubbelastung gefährdet Sensoren/Qualität. Sicherheitsrisiko auf Fahrwegen.",
"gains_add": "Staubfreie Umgebung ohne Produktionsstopp.",
"status": "Freigegeben",
"note_add": "Prio 1: Reinigung (Sweeper). Kein Stapler-Kampf!",
"ops_focus": True
},
"Logistics - Warehouse": {
"pains_add": "Staub auf Waren. Manuelles Kehren bindet Fachkräfte.",
"gains_add": "Werterhalt Hallenboden. Sauberkeit ohne Störung.",
"status": "Freigegeben",
"note_add": "Prio 1: Sweeper (Staub). Prio 2: Wet.",
"ops_focus": True
}
}
def run_updates():
print("--- Starting Notion Updates ---")
# 1. Update Personas
db_personas = find_db_id("Personas")
if db_personas:
print(f"Found Personas DB: {db_personas}")
for role, kpi_text in PERSONA_UPDATES.items():
page = get_page_id(db_personas, "Role", role) # Title col is "Role" here? Or "Name"? Script 1 said Role fallback.
# Actually, let's try "Name" first, then "Role".
# In previous dump, 'Name' was 'Infrastruktur-Verantwortlicher' etc.
# Let's assume the main column is "Name" (title).
if not page:
# Try finding by property "Role" (select) if title is different?
# Based on dump, the Title column content was "Infrastruktur-Verantwortlicher".
# Let's assume title property is named "Name" or "Role".
# Inspecting schema from previous run: `['Rollenbeschreibung', '...Product Categories', ... 'Role']`
# The Title property is likely "Role" or "Name".
# Let's try searching for "Role" property as title.
page = get_page_id(db_personas, "Role", role)
if page:
# Update KPIs
# Column name in schema: "KPIs"
update_page(page['id'], {
"KPIs": {"rich_text": [{"text": {"content": kpi_text}}]}
})
else:
print(f"⚠️ Persona '{role}' not found.")
else:
print("❌ Personas DB not found.")
# 2. Update Industries
db_ind = find_db_id("Industries")
if db_ind:
print(f"Found Industries DB: {db_ind}")
for vertical, data in INDUSTRY_UPDATES.items():
page = get_page_id(db_ind, "Vertical", vertical)
if page:
props = page['properties']
# Prepare updates
new_props = {}
# Status
# Check if Status is select or status
if "Freigabe" in props:
# Assuming Select or Status. Let's try Select first, if fails try Status
if props["Freigabe"]["type"] == "select":
new_props["Freigabe"] = {"select": {"name": data["status"]}}
elif props["Freigabe"]["type"] == "status":
new_props["Freigabe"] = {"status": {"name": data["status"]}}
# Ops Focus (Checkbox)
if "Ops Focus: Secondary" in props:
new_props["Ops Focus: Secondary"] = {"checkbox": data["ops_focus"]}
# Pains (Append)
current_pains = ""
if "Pains" in props and props["Pains"]["rich_text"]:
current_pains = props["Pains"]["rich_text"][0]["plain_text"]
new_pains = append_text(current_pains, data["pains_add"])
new_props["Pains"] = {"rich_text": [{"text": {"content": new_pains}}]}
# Gains (Append)
current_gains = ""
if "Gains" in props and props["Gains"]["rich_text"]:
current_gains = props["Gains"]["rich_text"][0]["plain_text"]
new_gains = append_text(current_gains, data["gains_add"])
new_props["Gains"] = {"rich_text": [{"text": {"content": new_gains}}]}
# Notes (Append)
current_notes = ""
if "Notes" in props and props["Notes"]["rich_text"]:
current_notes = props["Notes"]["rich_text"][0]["plain_text"]
new_notes = append_text(current_notes, data["note_add"])
new_props["Notes"] = {"rich_text": [{"text": {"content": new_notes}}]}
# Execute Update
update_page(page['id'], new_props)
else:
print(f"⚠️ Industry '{vertical}' not found.")
else:
print("❌ Industries DB not found.")
if __name__ == "__main__":
run_updates()

157
show_status.py Normal file
View File

@@ -0,0 +1,157 @@
import sqlite3
import os
import requests
import json
from datetime import datetime, timedelta
# --- Configuration ---
DB_PATH = "/home/node/clawd/repos/brancheneinstufung2/company_explorer_local.db"
CE_API_URL = "http://192.168.178.6:8090/ce/api"
# SO_API_URL = "..." # To be added
import sys
sys.path.append(os.path.join(os.path.dirname(os.path.abspath(__file__)), "connector-superoffice"))
from superoffice_client import SuperOfficeClient
class GtmHealthCheck:
def __init__(self):
self.db_path = DB_PATH
self.ce_api_url = CE_API_URL
self.api_user = os.getenv("COMPANY_EXPLORER_API_USER", "admin")
self.api_password = os.getenv("COMPANY_EXPLORER_API_PASSWORD", "gemini")
def get_ce_stats(self):
"""Holt Statistiken aus der lokalen Company Explorer DB."""
if not os.path.exists(self.db_path):
return {"total": 0, "in_progress": 0, "error": "DB file not found."}
try:
conn = sqlite3.connect(self.db_path)
cursor = conn.cursor()
cursor.execute("SELECT COUNT(*) FROM companies")
total = cursor.fetchone()[0]
cursor.execute("SELECT COUNT(*) FROM companies WHERE status != 'ENRICHED'")
in_progress = cursor.fetchone()[0]
conn.close()
return {"total": total, "in_progress": in_progress}
except Exception as e:
return {"total": 0, "in_progress": 0, "error": str(e)}
def check_ce_api_health(self):
"""Prüft die Erreichbarkeit der Company Explorer API."""
try:
response = requests.get(
f"{self.ce_api_url}/health",
auth=(self.api_user, self.api_password),
timeout=5
)
if response.status_code == 200 and response.json().get("status") == "ok":
return "[HEALTHY]"
return f"[ERROR - Status {response.status_code}]"
except requests.exceptions.Timeout:
return "[BUSY/TIMEOUT]"
except requests.exceptions.RequestException:
return "[UNREACHABLE]"
def get_so_stats(self):
"""Holt die ungefähre Gesamtanzahl der Firmen aus der SuperOffice API."""
try:
client = SuperOfficeClient()
# We query the first page with a page size of 200 (a common default)
query_string = "Contact?$top=200"
# Directly use the requests logic from the client's search method for a single page
url = f"{client.base_url}/{query_string}"
response = requests.get(url, headers=client.headers, timeout=15)
response.raise_for_status()
data = response.json()
count_on_page = len(data.get('value', []))
# Check for a next link to determine if there are more pages
if 'odata.nextLink' in data or 'next_page_url' in data:
return {"total": f"> {count_on_page}"} # More than one page
else:
return {"total": str(count_on_page)} # Exact number if only one page
except requests.exceptions.RequestException as e:
return {"total": f"API Error"}
except Exception as e:
return {"total": f"Error"}
def check_so_api_health(self):
"""Prüft die Erreichbarkeit der SuperOffice API."""
try:
client = SuperOfficeClient()
# A simple request to the base URL should suffice as a health check
response = requests.get(client.base_url, headers=client.headers, timeout=10)
if response.status_code == 200:
return "[HEALTHY]"
return f"[ERROR - Status {response.status_code}]"
except requests.exceptions.Timeout:
return "[BUSY/TIMEOUT]"
except requests.exceptions.RequestException:
return "[UNREACHABLE]"
def get_throughput(self):
"""Zählt die verarbeiteten Accounts der letzten Stunde aus dem Log."""
log_file = "/home/node/clawd/repos/brancheneinstufung2/logs/throughput.log"
if not os.path.exists(log_file):
return 0
count = 0
one_hour_ago = datetime.utcnow() - timedelta(hours=1)
try:
with open(log_file, "r") as f:
for line in f:
parts = line.strip().split(',')
if len(parts) >= 1:
try:
timestamp = datetime.fromisoformat(parts[0])
if timestamp >= one_hour_ago:
count += 1
except ValueError:
continue # Ignore malformed lines
return count
except Exception:
return "Log Error"
def render_dashboard(self):
"""Stellt das Dashboard auf der Konsole dar."""
ce_stats = self.get_ce_stats()
ce_health = self.check_ce_api_health()
so_stats = self.get_so_stats()
so_health = self.check_so_api_health()
throughput = self.get_throughput()
timestamp = datetime.now().strftime("%d.%m.%y %H:%M")
print("=======================================")
print(f"GTM Lead Engine - Status ({timestamp})")
print("=======================================")
print("\n[+] Schnittstellen:")
print(f" - SuperOffice API: {so_health}")
print(f" - Company Explorer: {ce_health}")
print("\n[+] Account Trichter:")
print(f" - SuperOffice Gesamt: {so_stats.get('total')}")
if 'error' in ce_stats:
print(f" - Im Company Explorer: Error ({ce_stats['error']})")
else:
print(f" - Im Company Explorer: {ce_stats.get('total')}")
print(f" - In Bearbeitung: {ce_stats.get('in_progress')}")
print("\n[+] Durchsatz (Letzte Stunde):")
print(f" - Verarbeitet: {throughput} Accounts")
print("\n")
if __name__ == "__main__":
checker = GtmHealthCheck()
checker.render_dashboard()