n8n Agent Workflow Visualization
// ENGINEERING LOG WORKFLOW_ID: AGENT_001

Automating the High Seas with n8n Agents

How I combined LangChain, n8n, and a bit of maritime grit to build autonomous agents that manage my creative workflow.

When you're out on the water, efficiency isn't just a luxury—it's survival. The same applies to managing a creative studio.

The Problem: Too Many Tabs, Not Enough Hands

As a boat captain turned AI engineer, my day is split between physical logistics and digital creation. I needed a way to:

Enter n8n: The Digital First Mate

I chose n8n because it's node-based (visual, like a chart) but allows for unlimited custom JavaScript (powerful, like a diesel engine).

The "Creative Agent" Workflow


// Example n8n Function Node: Prompt Enhancer
const mood = items[0].json.mood; // "Stormy Bayou"
const style = items[0].json.style; // "Cyberpunk"

return {
  json: {
    prompt: `Generate a high-fidelity image of ${mood} in a ${style} aesthetic. 
             Include neon lighting, water reflections, and 8k texture details.`,
    negative_prompt: "blurry, low resolution, distorted horizon"
  }
}
                

Connecting the Dots

Using webhooks, I connect these agents directly to my phone. I can text a concept like "Idea: Robot crab playing a saxophone", and the agent:

  1. Expands the prompt using an LLM.
  2. Generates 4 variations using Stable Diffusion.
  3. Saves them to my "Review" folder in Drive.
  4. Notifies me via Telegram when done.

What's Next?

I'm currently building a "Music Theory Agent" that analyzes my chord progressions and suggests bridge variations based on classic Blues scales. Stay tuned.

Back to Insights