app.polisplexity.tech/pxy_meta_pages/README.md

🧠 Core Purpose

The app links Facebook Pages to OpenAI assistants, listens to webhook events from Facebook (like comments or shares), and responds automatically with AI-generated replies. It logs all interactions for analytics or auditing.

---

🧩 Key Components and Their Roles

1. Models (models.py)

• FacebookPageAssistant: Links a Facebook Page (page_id, page_name) to an OpenAIAssistant, and tracks whether it's subscribed to webhooks.
• EventType: Catalog of event types like comment, share.
• FacebookEvent: Stores each received event, with sender, message, and timestamp.

2. Webhook Entry Point (views.py + urls.py)

• URL: /webhook/

• Accepts POST requests from Facebook.

• Uses item_type from the event payload to route to:

  • handle_comment_event
  • handle_share_event

3. Webhook Logic (webhook_handlers.py)

• verify_webhook_token: (Not currently used in the URL) for GET verification if needed.

• handle_comment_event:

  • Ignores self-comments.
  • Logs event in DB.
  • Uses OpenAI assistant to generate a reply to a comment.

• handle_share_event:

  • Logs share event.
  • Generates a thoughtful comment on the shared post (and optionally its parent).

4. Service Layer (services.py)

• FacebookService: Wraps Facebook Graph API calls.

  • Retrieves page access tokens.
  • Fetches post details (description, parent).
  • Posts comments or replies using OpenAI-generated responses.
  • Logs interaction in Neo4j via Neo4jDatabase.store_interaction(...).

---

🤖 AI Integration

• Uses OpenAIAssistant from pxy_openai to generate context-aware replies based on:

  • Original message or description of the post.
  • The assistant's configured personality or style.

• Bot personas like “Dr. Dr. Ekaropolus” are used to create engaging, scientifically-oriented content.

---

🔗 Data Flow

Facebook Webhook ──▶ /webhook/
                   └─▶ parse payload
                       └─▶ handle_comment_event()
                           ├─▶ store in DB
                           ├─▶ generate reply via OpenAI
                           └─▶ post to Facebook + log to Neo4j

---

🗂️ Persistence

• Uses Django ORM to store:

  • Page-assistant bindings.
  • Events (comments, shares).

• Logs interactions into a Neo4j graph DB for tracing conversations or influence paths.