Auto-Tag Sentiment Analyzer #4
Description
Auto-Tag Sentiment Analyzer
Fork:
https://github.com/CiaranOMahoney/intercom
Main repo:
https://github.com/Trac-Systems/intercom
Trac address:
trac1yxj0dk9w8wvjc3pw279xrkaettq5tpyy55f0zqhhhp4psruakf0qeqs007
Summary
Auto-Tag Sentiment Analyzer assigns a deterministic sentiment label to a conversation based on the customer’s first message. It outputs a single, human-readable tag (Positive or Negative) using fixed keyword rules.
How it works
Triggered when a new conversation message is received
Extracts the first customer message text
Applies rule-based keyword matching (no ML, no scoring)
Outputs one tag: Positive or Negative
Runs fully in demo mode without credentials
Intercom Integration
Operates on Intercom conversation message metadata
Uses deterministic rules only
Compatible with tagging and downstream workflows
Demo mode supported (no Intercom token required)
Proof
Server startup:
Request response:
