Feature Proposal: Native Gemini AI Semantic Email Classifier for GAS
Overview
Integrate a native Gemini AI LLM semantic classifier () into the suite to automatically categorize incoming unclassified emails and self-learn permanent Gmail filter rules.
Key Capabilities
- Zero-Configuration Semantic Classification: Classifies incoming emails from unknown senders based on intent, sender, subject, and body snippet into target canonical labels (, , etc.).
- Self-Learning Gmail Filter Generation: When Gemini classifies a new sender with high confidence (>= 95%), the script automatically calls to create a permanent Gmail filter rule.
- 100% Cloud Native: Executed natively inside Google Apps Script via calling or endpoints.
Proposed Architecture
- Runnable Apps Script module:
- Configuration schema:
- Test Coverage: Unit and integration Jest test suite in
Feature Proposal: Native Gemini AI Semantic Email Classifier for GAS
Overview
Integrate a native Gemini AI LLM semantic classifier () into the suite to automatically categorize incoming unclassified emails and self-learn permanent Gmail filter rules.
Key Capabilities
Proposed Architecture