Sample applications showcasing Auth0's authentication and authorization capabilities in the context of generative AI applications.
You can find a hosted demo here.
Samples are organized as below:
- authenticate-users: Easily implement login experiences, tailor made for AI agents. Whether for chatbots or background agents.
- call-apis-on-users-behalf: Use secure standards to get API tokens for Google, Github and more. Seamlessly integrate your app with other products.
- authorization-for-rag: Only retrieve documents users have access to. Avoid leaking data to a user that should not have access to it.
- asynchronous-authorization: Let your autonomous, async agents do work in the background. Use Async Auth to request approval when needed.
To support users' ability to download individual quickstarts from the Auth for AI Agents docs site, this repository contains a Github Action to generate downloadable zip file artifacts of each quickstart. This action will run on any change within a quickstart folder on main
branch, ensuring that the latest artifacts are always available for download.
To include a new quickstart, add a release-config.yml
file to the quickstart's base directory, for example:
// authorization-for-rag/langchain-js/release-config.yml
category: "authorization-for-rag"
framework: "langchain-js"
included: true
You can optionally add an exclude_patterns
to the release-config.yml
if you want to exclude any particular files from the release. Note that the release script already excludes and scans for common sensitive files. Example:
// release-config.yml
exclude_patterns:
- "*.tmp"
- "debug.log"
- "test-data/*"
- ".env.test"
- "node_modules/.cache/*"
To exclude an existing quickstart, within the release-config.yml
file, switch included
to false
.