Skip to main content

Concepts

Guardhouse sits between applications that need an identity and APIs that need to decide whether a request is allowed.

Client -- asks to sign in or obtain a token --> Guardhouse
Client <-- receives tokens ------------------- Guardhouse
Client -- sends an access token ------------> API resource
API resource -- validates and authorizes ----> Protected operation

The configuration has three layers:

  • Protocol: clients, API resources, scopes, grant types, and tokens
  • Identity: users and the authentication methods used to sign them in
  • Authorization: scopes, roles, permissions, and application policies

Standards and Discovery​

Guardhouse implements OAuth 2.0 and OpenID Connect endpoints under /connect/*. Applications should use the tenant issuer and its discovery document instead of hard-coding signing-key locations:

https://your-tenant.guardhouse.cloud/.well-known/openid-configuration

The discovery document identifies the issuer, supported capabilities, endpoint URLs, and JSON Web Key Set (JWKS) location.

The Security Boundary​

Guardhouse issues and validates identity credentials. The API receiving an access token must still:

  • validate the token rather than merely decode it
  • require the intended issuer and audience
  • reject expired or otherwise invalid tokens
  • enforce the access required by the requested operation

Continue with Protect an API for the implementation choices.