Deployer Training Materials Generator

Generate comprehensive, production-quality training documentation by scanning and analyzing a codebase using Gemini AI.

When to Use

• •Creating onboarding documentation for new developers
• •Generating deployer guides for operations teams
• •Understanding a new or unfamiliar codebase
• •Documenting a product before release
• •EU AI Act Articles 13 & 14 compliance (Transparency & Human Oversight)

Quick Start

python scripts/generate_training.py --path /path/to/repo

Instructions

1. •

   Identify the target codebase: Determine the repository root to scan.

2. •

   Run the generator:

   bash

   python "AI Act skills packages/AI Act package/deployer-training/scripts/generate_training.py" \
     --path <path-to-repository> \
     --name "Your Product Name"
   

3. •

   Review the output: Check project root Output/Deployer_Guide.md for the generated documentation.

4. •

   Human review: Always recommend human review for accuracy and completeness.

What You Get

A comprehensive Deployer_Guide.md with:

1. •Executive Summary: High-level product overview.
2. •System Architecture: Component diagrams (Mermaid), data flow, tech stack.
3. •Product Capabilities: Core features, user journeys, configuration options.
4. •Developer Onboarding: Environment setup, extension patterns, testing guidelines.
5. •Operational Guide: Deployment strategy, troubleshooting & limitations.

Parameters

Requirements

• •Python 3.8+
• •google-genai package: pip install google-genai
• •GEMINI_API_KEY environment variable set

Best Practices

• •Run on a clean checkout of the repository.
• •The generator respects .gitignore and excludes .env files for security.
• •For large codebases, review the "Total context size" output to ensure it fits in the model's context window.
• •Always perform human review on generated documentation.

EU AI Act Compliance

This tool addresses:

• •Article 13: Transparency and provision of information to users
• •Article 14: Human oversight requirements (documentation for deployers)