Raised: $0
0% of monthly goal Help us cross the finish line!
Goal: $12,000
Raised: $0 Goal: $12,000
0% of monthly goal Help us cross the finish line!
Sponsor DDEV

add-on registry tests last commit release

DDEV AI Agent

Overview

A DDEV add-on to spin up a simple, opinionated, customizable n8n/supabase AI Agent workflow stack with sensible defaults.

Getting Started

Create a DDEV project.

mkdir my-ai-agent
cd my-ai-agent
ddev config --project-name=my-ai-agent --project-type=generic --docroot=.

Then, get this add-on and start the project.

ddev add-on get jcandan/ddev-ai-agent
ddev restart
ddev launch :5678

After launch, you may commit the .ddev directory to version control.

Requirements

Get Started

Install and start using DDEV.

NOTE: DDEV touts itself as a Docker-based PHP development environment, but it really is a docker-compose abstraction that affords us the ability to simplify providing you with a customizable AI Agent development environment with sensible, opinionated defaults. If you’re not interested in PHP, you can proceed to use this project undaunted.

Usage

Command Description
ddev start / ddev restart Launch or rebuild the stack
ddev describe View services and URLs
ddev logs -s n8n Tail n8n logs
ddev logs -s supabase-db Tail Postgres logs

Access n8n at https://<project>.ddev.site:5678

Optional local LLM (Ollama)

This add-on can optionally start a local Ollama LLM service using a DDEV profile.

Start the stack with the ollama profile:

ddev start --profiles ollama

From other services in the stack (for example, n8n), you can reach the local LLM at:

http://ollama:11434

Use that URL as the base for your Ollama / OpenAI-compatible nodes in n8n when the profile is enabled.

You can check the service connection with:

ddev exec curl http://ollama:11434/api/version

The service will be picked up by subsequent ddev restart, but to stop the ollama service, run ddev stop and ddev start again without the profile.

Roadmap

See Roadmap for details about our plans for the future.

Credits

Contributed and maintained by @jcandan

Inspired by @coleam00’s local-ai-packaged.