DrupalAI Hackathon 2026

This project is based on Drupal CMS, a fast-moving open source product that enables site builders to create Drupal sites with smart defaults using only their browser.

This repository serves as the official starting point for hackathon participants.

Table of Contents

• Participation workflow
• Getting started
• AI provider and model requirements
• amazee.io AI provider setup
• DevPanel setup guide
• Submitting your work
• Resources

Participation workflow

Participants must fork this repository into a GitHub account of their choosing.

All development must happen in the forked repository.

When ready for review, participants must open a pull request back to this repository so that judges can review and evaluate the submission.

Getting started

If you want to use DDEV to run Drupal CMS locally, follow these steps:

1. Install DDEV following the documentation at https://ddev.com/get-started/
2. Open a terminal and change to the root directory of this project
3. Run the following commands:

ddev start
ddev install
ddev launch

Drupal CMS has the same system requirements as Drupal core. You may use any supported local setup if you prefer.

See the Drupal User Guide for more information on installation options: https://www.drupal.org/docs/user_guide/en/installation-chapter.html

AI provider and model requirements

This distribution is configured to use an AI provider via the Drupal AI module. Participants must use the amazee.io provider for all AI-powered features in their submission. Non-amazee.ai providers are not permitted unless explicitly approved by hackathon organizers.

Participants may add any Drupal modules they require for their solution.

amazee.io AI provider setup

To configure the provider:

1. Visit the amazee.io provider configuration form at Configuration → AI → Provider Settings → amazee.io Authentication (typically /admin/config/ai/providers/amazeeio).
2. Enter the email address used to register for the hackathon to start the authentication process. Please note that email matching is case-sensitive on the backend, so be sure to use the exact casing you registered with (e.g., John.SMITH@example.com rather than john.smith@example.com).
3. Check your email for a verification code from amazee.ai, enter it in the form, and submit.
4. Choose the region Play to Impact - Drupal AI Hackathon 2026. If you don't find it contact the organizers.
5. Once verified, the amazee.ai credentials (LLM key and VectorDB key) will be stored in the Keys module at /admin/config/system/keys.
6. Go to Configuration → AI → AI Default Settings and select at least one Mistral model at its core for the relevant operation types.

Please note that the keys generated by amazee.io are exported in clear text in the configuration. For this reason, the related configuration has been config-ignored. As a result, when setting up the site on DevPanel, you must configure the keys again by following the instructions above.

DevPanel Setup Guide

This section explains how to set up your DevPanel project and development environment for the hackathon.

Although not mandatory, teams are strongly encouraged to use DevPanel so that non-technical team members and jury members can easily access and test the project.

Roles and Responsibilities

• Each team member may create their own individual DevPanel account.
• Only one DevPanel site is required per team.
• Each team has a designated tech lead, who will be responsible for:
  • Creating and deploying the DevPanel project
  • Keeping the site up to date throughout the hackathon
  • Communicating required details to the hackathon organizing committee
  • Managing access to the DevPanel workspace
• The tech lead can grant access to all other team members, allowing everyone on the team to view and interact with the site.

1. Setting Up a DevPanel Account

To use DevPanel, each team member must create an individual DevPanel account using the same email address they used for hackathon registration.

After registering, the tech lead must send the following information to the hackathon organizing committee on behalf of the team:

• Email addresses used for DevPanel registration
• Full names of team members
• Team [number] - [team name]

2. Forking the Source Repository on GitHub

Steps (performed by the tech lead):

1. Go to the official hackathon source repository.
2. Create a fork under your personal GitHub account.
3. Do not rename the repository.

Teams are free to collaborate directly on the fork or in any other ways they see fit.

Important: The forked repository is your team’s single source of truth. DevPanel environments are disposable, but your GitHub repository is not.

3. Creating a DevPanel Project from the Forked Repository

After creating a DevPanel account, the tech lead should complete the following steps:

1. In the Workspaces section, locate the Hackathon organization workspace your team has been added to and click it to enter.
2. Click + Create Project from Scratch.
3. In the project creation form, configure the project as follows:
   • Name: Team [number] - [team name]
   • Application Type: Drupal 11
   • Source Code: GitHub → Link your account by signing in to GitHub
   • Repository: Choose Use existing repository, then select the forked repository created in Step 2
   • Branch: Select master
4. Confirm and create the project.
5. Go to the Branches tab and click master.
6. Configure the environment:
   • Database & Static Files Settings: From scratch
   • Security Settings: Use default settings
   • Configuration Settings: Use default settings
   • Review and Launch: Click Deploy
7. Wait for the project to finish building (this may take a few minutes).

Once complete, the Application Summary page will be available. The site is automatically built and configured using the .devpanel folder in the reference repository.

4. Granting Access to Team Members

After deployment, the tech lead must add the remaining team members to the DevPanel workspace using the Members tab. This allows all team members to access, review, and demonstrate the site.

5. Build and Install the Site

This repository includes a .devpanel directory containing all the required configuration for DevPanel to correctly build and install the site.

Once the setup is complete, select your fork’s default branch (most likely master) and deploy it by following the on-screen instructions.

After the deployment is complete, you can run any Drush commands on your site (for example, drush uli) by following the steps below:

1. From the Application Summary page, click Open Application to access the development environment (browser VS Code).
2. Copy the displayed password — you will need it to access the environment.
3. In browser VS Code open a terminal window and run the desired commands.

Important: You must always set up the amazee.io AI provider manually on the DevPanel site. Its configuration is intentionally ignored and will not be exported with drush cex -y. For more information, see the Amazee.io AI Provider Setup section above.

6. Troubleshooting

If the site does not build and install automatically during deployment, you can perform the process manually by following the steps below in a VS Code terminal window:

composer install
drush -y si --existing-config --db-url="${DB_DRIVER}://${DB_USER}:${DB_PASSWORD}@${DB_HOST}:${DB_PORT}/${DB_NAME}"

In the event of temporary cluster unavailability, you may see the following message when attempting to access either the project URL or the VS Code URL:

{
  "message": "An invalid response was received from the upstream server"
}

If this occurs, wait a few minutes and refresh the page to check whether the issue has been resolved.

Submitting your work

By the start of the pitch presentation, all work done in the fork repository must be submitted via pull request to the reference repository.

Make sure to commit and include any configurations by running the following command prior to committing:

ddev drush cex

Note: Only 1 PR per team will be accepted.

Resources

While not included in this distribution, the following modules may be useful resources to explore during the hackathon if they align with your project goals:

• FlowDrop UI Agents: A visual agent builder that greatly improves the development experience for creating AI-powered agents.
• Tool API: A module that allows you to define and manage tools that can be used by AI agents to perform specific tasks. This represents a more integrated approach for agents to interact with tools.
• Agent Handler: When combined with FlowDrop UI for agents, this provides an AI agent that can create and edit agents and explain available tools. You can attach it to a chatbot to get assistance with agents and tools.
• Pexels AI: Integrates the Pexels API to provide AI-powered image search and download capabilities.