DDD Agent

An AI-powered agent that generates LinkedIn posts about Domain-Driven Design (DDD) topics, built with Embabel Agent Framework.

Quick Start

export GOOGLE_STUDIO_API_KEY=your-google-studio-key
ollama pull llama3.2
mvn spring-boot:run

The app launches an interactive shell. Type x "your topic" and the agent will produce a reviewed, publication-ready LinkedIn post saved to posts/.

Prerequisites

• Java 25+
• Maven 3.9+
• Ollama running locally with a model pulled
• A Google AI Studio API key → aistudio.google.com (free tier works)

How It Works

The agent runs a two-step pipeline powered by Embabel, a JVM agent framework that uses Goal-Oriented Action Planning (GOAP) to resolve the execution order automatically from input/output types — no orchestration code needed.

UserInput ──► writeDraft (Ollama) ──► reviewPost (Gemini) ──► LinkedInReviewedPost

1. writeDraft — A local Ollama model writes an opinionated draft from the perspective of a developer who has applied DDD in real systems.
2. reviewPost — Gemini verifies DDD terminology accuracy and LinkedIn writing quality (hook, mobile readability, takeaway), then saves the final post to posts/.

Usage

Once the shell starts, use the x command with your DDD topic:

x "bounded contexts explained"