* Move the rest of TestingResource methods to a utils-shared helper class (1/2)
Closes: #48708
Signed-off-by: Simon Vacek <simonvacky@email.cz>
* rename runOnServer field
Signed-off-by: Simon Vacek <simonvacky@email.cz>
* remove realm name from method input
Signed-off-by: Simon Vacek <simonvacky@email.cz>
---------
Signed-off-by: Simon Vacek <simonvacky@email.cz>
Adds Shared Signals Framework support to Keycloak in the **SSF Transmitter** role: Keycloak signs Security Event Tokens (SETs, RFC 8417) describing realm/user/session/credential events and delivers them to OAuth clients
registered as **SSF Receivers**, either by HTTP PUSH (RFC 8935) or HTTP POLL (RFC 8936).
Targets the OpenID Shared Signals Framework 1.0 (Final) specification plus the CAEP Interoperability Profile 1.0. Ships the legacy SSE CAEP profile alongside for Apple Business Manager / Apple School Manager interop, since Apple device-fleet enrolment is a concrete drive-use case.
Gated behind \`Profile.Feature.SSF\` experimental, opt-in.
Issue #43614 originally proposed SSF *Receiver* support (Keycloak ingesting SETs from upstream IdPs / risk engines). After exploring both sides, we're shipping the **Transmitter** first (see #48254) because it covers the strongest community asks (federate Keycloak events to downstream SaaS, Apple device fleet revoke flow) and lets us validate the SSF data-plane against real receivers before designing the harder "action mapping" question on the Receiver side. Receiver support remains on the roadmap and is tracked separately via #43614.
**In:**
- Compliance with SSF 1.0, CAEP 1.0, RISC 1.0, RFC 8935, RFC 8936, RFC 9493, RFC 8417
- SSF Transmitter support (Keycloak Realm can act as a SSF Transmitter)
- SSF Stream management (CRUD, status, verification)
- SSF Subjects management (subjects)
- SET delivery via HTTP PUSH (RFC 8935) and HTTP POLL (RFC 8936) with POLL in a return-immediately form
- SSF events temporarily stored in durable outbox with cluster-aware drainer and exponential backoff
- SSF Receivers managed as OIDC Clients with client credentials grant or auth code grant (currently only one stream per client)
- Support for SSF Stream, CAEP 1.0 and RISC 1.0 events (custom events via SPI)
- CAEP credential-change / session-revoked / (device-compliance-change) event mapping from native Keycloak events
- Support for RFC 9493 Subject Identifiers for Security Event Tokens
- Support for SSF Receiver subject event subscription with subject selection (per-user / per-orgssf.notify.<clientId>attribute, support fordefault_subjectspolicy (ALL, NONE))
- Support for Synthetic event emittance via REST endpoint for non-Keycloak-native event sources (external IAM solution)
- Per-receiver "Emit-only events" gate to suppress auto-emit per event type per receiver
- Support for legacy SSE CAEP profile for Apple Business Manager / Apple School Manager interop (verified)
- Per-realm SSF admin REST + Admin UI for SSF-enabled clients (Receiver / Stream / Subjects / Events)
- Prometheus metrics (dispatcher, drainer, poll, verification, outbox depth, delivery metrics)
**Out (tracked as separate follow-up issues):**
- SSF Receiver role for Keycloak (ingestion of SETs)
- POLL long-polling (\`returnImmediately=false\` honoured)
- Dedicated SSF signing key (separate from realm OIDC signing key)
- Chunked HELD release for very large backlogs
- Performance characterization + security review
- Formal interop matrix (caep.dev, ABM)
- [X] All code gated behind \`Profile.Feature.SSF\` (experimental, off by default)
- [X] Per-realm \`ssf.transmitterEnabled\` toggle; per-client \`ssf.enabled\` toggle
- [X] SSF event listener registered as global (not user-toggleable per realm)
- [X] Receiver-facing endpoints conformant with SSF 1.0
- [X] CAEP credential-change / session-revoked / device-compliance mapping pass interop testing against \`caep.dev\`
- [X] SSE CAEP profile narrowed shape works with Apple Business Manager
- [X] Integration test coverage for the dispatch / outbox / push / poll pipeline (100+ tests)
- [X] Prometheus metrics exposed under \`keycloak_ssf_*\`
- [X] Design notes published
Fixes#48901
This PR was partially co-authored with Claude AI
Signed-off-by: Thomas Darimont <thomas.darimont@googlemail.com>
possible values: conditional, optional, required, silent
conditional remains the default to not break the current behavior
when optional or required and the user dismissed the modal, it will stay hidden for this auth-session, can still be opened by button
adjusted all related resources, like JS files (also consolidated duplicated logic), Java classes and freemarker template
tests extended
passkey documentation extended/updated
closes#46959
Signed-off-by: Niko Köbler <niko@n-k.de>
Closes#47860
Signed-off-by: Pedro Ruivo <1492066+pruivo@users.noreply.github.com>
Co-authored-by: Pedro Ruivo <1492066+pruivo@users.noreply.github.com>
Closes#47187
Signed-off-by: Pedro Ruivo <1492066+pruivo@users.noreply.github.com>
Signed-off-by: Alexander Schwartz <alexander.schwartz@ibm.com>
Signed-off-by: Pedro Ruivo <pruivo@users.noreply.github.com>
Co-authored-by: Pedro Ruivo <1492066+pruivo@users.noreply.github.com>
Co-authored-by: Alexander Schwartz <alexander.schwartz@ibm.com>
Co-authored-by: Steven Hawkins <shawkins@redhat.com>
* Fix @TestSetup method ordering with inheritance
ReflectionUtils.listMethods() had two bugs when class inheritance was
involved:
1. Subclass methods were added before superclass methods, resulting in
wrong execution order. Expected behavior is superclass first, matching
standard Java inheritance semantics.
2. When a subclass overrides a superclass method annotated with
@TestSetup, both methods were collected causing the subclass method
to execute twice instead of only the override running once.
The fix builds the class hierarchy top-down (superclass first) and skips
superclass methods that are overridden by a subclass method with the
same name and parameter types.
Closes#46667
Signed-off-by: hammadxcm <hammadkhanxcm@gmail.com>
* Small refactoring to add when not already added, instead of adding/remove
Signed-off-by: stianst <stianst@gmail.com>
---------
Signed-off-by: hammadxcm <hammadkhanxcm@gmail.com>
Signed-off-by: stianst <stianst@gmail.com>
Co-authored-by: stianst <stianst@gmail.com>
Closes#46557
Signed-off-by: Pedro Ruivo <1492066+pruivo@users.noreply.github.com>
Signed-off-by: Alexander Schwartz <alexander.schwartz@ibm.com>
Co-authored-by: Pedro Ruivo <1492066+pruivo@users.noreply.github.com>
Co-authored-by: Alexander Schwartz <alexander.schwartz@ibm.com>
