Skip to content

AI Features

AI-powered features that enhance the documentation experience, built with modern cloud services and a focus on user privacy.

  • Chat Widget

    A Claude Docs-inspired AI chat widget that provides instant, conversational answers about the site content.

    Chat Widget Overview

  • Architecture

    MVVM frontend pattern with modular backend. Google ADK multi-agent orchestration with session memory.

    Architecture Guide

  • Deployment

    Deploy to Google Cloud Run with uv for fast builds. Modular agent structure for easy customization.

    Deployment Guide

  • Integration

    Plug-and-play guide for adding the chat widget to your own Zensical or MkDocs Material site.

    Integration Guide

Current AI Features

Chat Widget

The primary AI feature is an interactive chat widget that:

  • Answers questions about site content, navigation, and projects
  • Remembers context within a conversation session
  • Works on all devices with responsive mobile/desktop layouts
  • Respects privacy with no conversation logging

Status: Production-ready and deployed (v3.0)

Technology Stack:

Layer Technology
Frontend JavaScript MVVM, DOM injection
Backend Flask + Google ADK on Cloud Run
AI Model Google Gemini 2.5 Flash
Orchestration Multi-agent with 7 specialized sub-agents
Security Rate limiting, CORS, prompt injection detection

Roadmap

Planned AI features:

  • [ ] Smart Search - AI-enhanced search with semantic understanding
  • [ ] Content Summaries - Auto-generated page summaries
  • [ ] Code Explanations - Inline explanations for code snippets
  • [ ] Multi-language Support - Chat in multiple languages