Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs-chat.die-coaching-akademie.de/llms.txt

Use this file to discover all available pages before exploring further.

The app uses two JSON knowledge sources:
FilePurpose
data/knowledge_base.jsonCurated, stable answer basis
data/website_knowledge_base.jsonGenerated website-content entries

​
Entry shape

Curated entries follow this shape: 
{
  "id": "beispiel-thema",
  "category": "Ausbildung",
  "title": "Beispiel-Thema",
  "summary": "Kurzer Such- und Prompt-Kontext.",
  "keywords": ["beispiel", "synonym"],
  "content": "Gepruefte Antwortgrundlage ohne erfundene Details.",
  "next_step": "Empfiehl die passende Angebotsseite.",
  "links": [
    {
      "label": "Mehr erfahren",
      "url": "https://www.die-coaching-akademie.de"
    }
  ],
  "priority": 3
}

​
Selection logic

select_relevant_entries scores entries using:
  • keyword matches
  • query token overlap with title, summary, and content
  • entry priority
  • special boosts for consulting and leadership queries

​
Website sync

Website source definitions live in: 
data/website_sources.json
Run the sync: 
cd faq_bot
source .venv/bin/activate
python sync_website_content.py
On production, restart after sync so the loaded knowledge base refreshes: 
systemctl restart chatbot
Dates, prices, and availability should stay grounded in current website or administrative sources. Avoid hard-coding uncertain details unless they are explicitly verified.