AdminForth Adapters
This page has a list of all available adapters in the AdminForth ecosystem. Adapters are used by AdminForth plugins to implement connections to various 3rd-party services and APIs. Every adapter could be easily forked and customized for any new service or API.
📧 Email Adapters
Used to send emails
AWS SES Email Adapter
npm i @adminforth/email-adapter-aws-ses
Enables email delivery via Amazon Simple Email Service (SES), suitable for high-volume, programmatic email sending.
Mailgun Email Adapter
npm i @adminforth/email-adapter-mailgun
Allows sending transactional or marketing emails using Mailgun, a developer-friendly email service.
🔐 OAuth2 Adapters
Used to authenticate users via OAuth 2.0 providers
Google OAuth Adapter
npm i @adminforth/google-oauth-adapter
Supports Google sign-in to allow users to authenticate using their Google or Google Workspaces accounts.
GitHub OAuth Adapter
npm i @adminforth/github-oauth-adapter
Enables authentication via GitHub accounts, useful for developer tools and open-source apps.
Facebook OAuth Adapter
npm i @adminforth/facebook-oauth-adapter
Allows users to log in with Facebook credentials. Facebook OAuth is commonly used for social media integrations.
Keycloak OAuth Adapter
npm i @adminforth/keycloak-oauth-adapter
Connects AdminForth to an Open-Source Keycloak - generally self-hosted - identity provider for enterprise-grade SSO (Single Sign-On).
Microsoft OAuth Adapter
npm i @adminforth/microsoft-oauth-adapter
Supports login through Microsoft accounts including Azure AD, Office365, and Outlook.com.
Twitch OAuth Adapter
npm i @adminforth/twitch-oauth-adapter
Adds support for Twitch authentication, useful for streaming or creator-oriented platforms.
🎨 Image Generation Adapters
Is used for image generating AI tools.
🔗ImageGenerationAdapter source class
OpenAI Image Generation Adapter
npm i @adminforth/image-generation-adapter-openai
Uses OpenAI’s image generation models (like DALL·E and gpt-image-1) to generate images from text prompts.
Up to the summer 2025 OpenAI models are one of the most powerful image generation models available (Especially GPT-Image-1), that is why we started with them.
💾 Storage Adapters
Is used for storing files
Amazon S3 Storage Adapter
npm i @adminforth/storage-adapter-amazon-s3
Stores uploaded files in Amazon S3, providing scalable cloud storage. Can be easily forked and customized to work with any S3-compatible storage service like MinIO or Wasabi or 3rd-party S3-capable services.
Local Storage Adapter
npm i @adminforth/storage-adapter-local
Stores files locally on the server’s filesystem. Suitable for development or small-scale self-hosted setups. Not really recommended for production use, cause cloud storage is more reliable and scalable.
🧠 AI Completion Adapters
🔗CompletionAdapter source class
Is used for AI-powered text completion. Feel free to fork and implement other models including models from Anthropic, Google Gemini, or any other AI model that supports text completion.
OpenAI Completion Adapter
npm i @adminforth/completion-adapter-open-ai-chat-gpt
Integrates AdminForth with OpenAI’s ChatGPT models to provide AI-powered completion and conversational features.
🔎 Image Analysis
🔗ImageVisionAdapter source class
Used for AI-powered image analysis. These adapters can automatically analyze image content, extract text from images, identify objects, describe scenes, and provide detailed image insights.
OpenAI Vision Adapter
npm install @adminforth/image-vision-adapter-openai --save
Integrates AdminForth with OpenAI to provide advanced AI-powered image analysis. Can describe image content, read and extract text from images, identify objects and people, and provide detailed visual insights.
🗄️Key-value Adapters
🔗Key-value adapter source class
Key-value adapters are used to store data in a key-value format. They provide a simple and efficient way to manage data where quick access to values based on unique keys is required.
RAM Adapter
npm i @adminforth/key-value-adapter-ram
The RAM adapter is a simplest in-memory key-value storage. Stores data in process RAM memory. Not sutable if you run application with several processes because each process will have own RAM. In last case you need centralized KV adapter, e.g. Redis.
Pros:
- Simplest in use - does not reqauire any external daemon. Cones:
- In production sutable for single-process installations only