Canterbury

The Canterbury release expands Strivacity with delegated authorization, mobile authentication, richer authentication context, and broader risk-based access controls.

This release includes changes that may impact your Strivacity deployment. See Important notes for details.


New features and enhancements

Strivacity for Agentic AI

This release introduces significant improvements to the initial Strivacity for Agentic AI capabilities supporting delegated authorization, enabling AI agents and other applications to request access to protected MCP servers and APIs on behalf of customers.

Key capabilities include:

  • Configure MCP servers and APIs as protected resources for delegated access
  • Assign protected resources to OAuth/OIDC clients and control which scopes each client can request
  • Configure mandatory and optional scopes for customer authorization consent
  • Use protected resource URIs as audiences for MCP clients in accordance with RFC 8707
  • Present customers with authorization consent when an agent or application requests access to protected resources
  • Let customers review and revoke authorization consents through self-service account management

Push notification MFA

This release introduces improved capability for push notifications as an MFA method, allowing customers to verify authentication requests from your brand's mobile application.

Key capabilities include:

  • Configure push notifications for Android and iOS applications, using Firebase Cloud Messaging (FCM) for Android and iOS or Apple Push Notification service (APNs) for iOS
  • Support Approve or Deny and Approve with number matching authentication methods
  • Enable device authenticator MFA in Adaptive Access policies and configure the approval lifetime
  • Let customers enroll multiple mobile devices and manage enrolled devices in My Account
  • Monitor push notification activity and authentication outcomes

Adaptive Access for token lifecycle events

Adaptive Access blocking rules now apply to token lifecycle events, extending risk-based enforcement beyond interactive authentication journeys.

IP address and location blocking rules can deny token issuance, refresh, exchange, and delegation requests when a request matches a configured rule.

Key capabilities include:

  • Apply existing Adaptive Access blocking rules to authorization code, refresh token, client credentials, and token exchange requests
  • Block token requests based on configured IP address and location rules
  • View Adaptive Access evaluation results in account events
  • Continue token processing normally when no blocking rule matches

Login anomaly notifications

Strivacity can now notify customers after successful logins from new devices or new locations, helping customers identify potentially suspicious account activity.

Key capabilities include:

  • Detect successful logins from devices and locations that Strivacity does not recognize for the customer
  • Configure new-device and new-location notifications separately in Adaptive Access policies
  • Set new-location detection sensitivity to Country, Region, or City
  • Let customers manage corresponding notification preferences in My Account
  • Customize notification templates
  • Review delivery activity in Notification history

Authentication context claims

Strivacity now provides authentication context in ID tokens through the amr and acr claims, allowing OIDC clients to evaluate the authentication methods used and the assurance level achieved.

Key capabilities include:

  • Include amr and acr in ID tokens based on the current authentication event
  • Represent the authentication methods used through amr, including password, OTP, FIDO/WebAuthn, push notification MFA, and multi-factor authentication
  • Represent the achieved NIST authentication assurance level (AAL1, AAL2, or AAL3) through acr
  • Use FIDO/WebAuthn authenticator metadata when determining authentication method and assurance
  • Configure application clients to include ACR only, or ACR and AMR, in access tokens

Risk policies

Risk policies provide a reusable way to evaluate multiple risk signals and produce a result that can be used during customer journeys and Adaptive Access evaluation.

Key capabilities include:

  • Configure reusable risk signal sources for geolocation and phone risk evaluation
  • Define custom risk results and evaluate signals using ordered rules
  • Calculate risk scores by assigning values to signal levels and mapping score ranges to results
  • Evaluate a Risk policy directly in a customer journey and route the customer based on the result
  • Use Risk policy results in Adaptive Access to determine the appropriate authentication action

Other enhancements

  • The After token refresh hook now provides the originating IP address and resolved location data for token refresh requests.
  • Dashboard filtering now supports Identity store and Organization filters, allowing administrators to view metrics for a selected identity store or one or more organizations within it. Organization filtering applies only to data collected after the feature was deployed.
  • Customers now receive a clear validation error when a password contains their name, username, or other restricted words.
  • Customers now receive a clear validation error when a password contains their name, username, or other restricted words.
  • API callers can specify whether possession of an email or phone identifier has already been verified when creating the identifier.
  • Account listings in the Admin Console now show the organizations associated with each account.
  • Locale support now includes English (Ireland) (en-IE) and Norwegian (Norway) (no-NO).
  • RP-initiated logout now supports POST requests.
  • Customer and admin account APIs can now set the language assigned to an account.

Bug fixes

We fixed issues where:

  • Username reminder SMS messages could fail when using a custom hook telephony provider.
  • Invalid Accept-Language headers could cause translation requests to fail.
  • The Forward customer to external provider setting in organization policies required an extra save before it could be enabled.
  • Lifecycle event hook error notifications could appear empty after correcting and saving a syntax error.
  • Before ID token generation hook failures could result in an internal server error.
  • Token request failures could interrupt notification and event streaming delivery.
  • OAuth2/OIDC Native client logout URL validation could prevent corrected configurations from being saved.
  • Bulk operations could fail with authorization errors when using client credentials authentication.
  • Login Gateway clients allowed unsupported token endpoint authentication methods to be selected.
  • Bulk operations could incorrectly process malformed input or return incorrect validation errors for requests exceeding limits.
  • Passkey enrollment and authentication requests with missing credential data could result in an unexpected error.
  • Organization Portal administrators could not add authenticators to disabled accounts.

In Build 1, we fixed issues where:

  • The brand logo could be missing from Admin Console login error pages.
  • Boolean and date values from external login providers could fail to map correctly.
  • Admin API requests using client credentials with JWT access tokens could fail when retrieving identity store details.
  • An updated account name could remain unchanged in parts of the Admin Console until the page was reloaded.
  • Restricted-character guidance could display incorrectly in the password change dialog.

Important notes

This release contains potential breaking changes for some customers:

  • The After token refresh hook now provides request data through the request object. Existing hooks that use request_body continue to work temporarily, but should be updated to use request.body.
  • Lifecycle Event Hooks now run on Node.js 22 instead of Node.js 20. Review custom hook code and dependencies for compatibility with Node.js 22.

Deprecations

OAuth 2.0 implicit grant flow

The OAuth 2.0 implicit grant flow is now deprecated. This change supports ongoing improvements to platform security and aligns with current OAuth security best practices.

This capability will remain available in the Canterbury release and will be fully removed in the next release.

Customers currently using the implicit grant flow should contact Strivacity Customer Success immediately to plan their migration to Authorization Code with PKCE.