Introduction
The AI Doc Generator is a multi-agent documentation system that uses Claude Code to orchestrate the generation of professional LaTeX documents from source materials.
System Overview
Key Concepts
Multi-Agent Architecture
The system uses a hierarchical agent structure:
| Agent | Model | Role |
|---|---|---|
| Orchestrator | Claude Opus | Controls workflow, makes decisions, merges context |
| Workers | Claude Sonnet | Execute parallel tasks (reading, writing, reviewing) |
Parallel Processing
Multiple agents work simultaneously:
- Source documents read in parallel
- Chapters written concurrently
- Reviews run asynchronously
External Research Integration
The orchestrator generates prompts for external AI tools:
- NotebookLM - Synthesize uploaded source documents
- Perplexity - Gather current statistics and trends
- Scopus AI - Find academic papers with BibTeX citations
Workflow Phases
- Source Collection - Gather PDFs and documents into
sources/ - Parallel Analysis - Sonnet agents read sources simultaneously
- External Research - Run prompts on NotebookLM, Perplexity, Scopus
- Document Planning - Opus plans structure with chapter assignments
- Parallel Writing - Writer agents create chapters concurrently
- Review & Build - Review agents check quality, LaTeX compiles to PDF
Technology Stack
AI Tools
| Tool | Purpose |
|---|---|
| Claude Code | Primary orchestration |
| OpenCode | Alternative multi-provider CLI |
| NotebookLM | Source document synthesis |
| Perplexity | Current data research |
| Scopus AI | Academic paper search |
Output Technology
| Component | Technology |
|---|---|
| Document Format | LaTeX with BibLaTeX |
| Build System | pdflatex + biber |
| Diagrams | TikZ |
| Code Highlighting | listings |
Getting Started
Ready to start?
- Quick Start - Set up in 10 minutes
- Full Workflow - Detailed process guide
- Agent Prompts - Ready-to-use prompts