ARCHIVED CHAPTER 09

Integration and Application Development

Adopt individual capabilities, connect existing applications, and build branded experiences through defined interfaces.

Organizations rarely replace their entire technology estate to adopt one new capability. Hybrid-Chain's integration model is designed around selective adoption: an existing application can connect to a relevant module, and a broader platform can compose several modules into a branded experience.

Contracts as the integration boundary

The public API reference organizes capabilities into modules and operations. For a working integration, the target deployment's current contract defines the request, response, authorization, and error behavior. Product descriptions explain purpose; machine-readable contracts and operation documentation establish the interface an engineer can actually use.

An operation's presence in a catalog should be read together with its status. A planned capability describes intended scope, while a published executable contract describes an implemented interface. Even an implemented interface can depend on permissions, resource state, configuration, or downstream availability. This distinction lets teams plan without mistaking a roadmap entry for an enabled service.

From request to confirmed outcome

Reliable integration requires understanding what a response means. Acceptance may indicate that work has been received, not that it has finished. Event delivery may inform an application that something changed, while a subsequent authorized read establishes the current state. Retries should follow the documented operation semantics and reconcile uncertain results before creating a new consequential request.

Automation and notification capabilities help connect platform events to external applications. The receiving application remains responsible for verifying the relevant message and authorizing its own side effects. A delivery acknowledgment and a completed business process are separate outcomes, and operational interfaces should make both understandable.

Choosing an integration surface

ApproachBest suited toEngineering responsibility
REST / documented request-response APIExplicit reads, proposals, and individually authorized operations.Validate the exact contract, permissions, idempotency behavior, and meaning of acceptance.
Events and real-time updatesKeeping a UI or downstream workflow informed when relevant state changes.Use the supported transport; establish authentication, ordering, replay, deduplication, and a canonical read for reconciliation.
White-label or embedded applicationA branded journey that combines several product capabilities.Preserve custody, permission, privacy, and availability boundaries behind the presentation.

These approaches are complementary. An embedded merchant interface may issue a documented request, receive a supported update, and then read authoritative state before showing completion. WebSockets are one real-time interaction pattern; they should not be assumed to be available for every module. Select the transport and delivery guarantees actually documented for the target deployment.

Branded and embedded experiences

White-label and partner applications can present a business-specific journey while drawing on the platform's capabilities. The value is the ability to focus the user experience on a particular market or operating process without independently assembling every underlying module. Branding does not change the actual authority, custody, or availability boundaries of the chosen services.

An integration can start with an authorized read, evidence inspection, or policy evaluation before adding an enabled action. This provides a practical progression: establish the meaning of the data, agree responsibilities, test expected and denied cases, and then evaluate the complete workflow under its intended conditions.

Developer value

The platform's modular structure reduces the need to invent a separate conceptual model for each connected task. Identity, resource context, authority, and evidence recur across products. Engineers still implement the exact contracts they use, but they can reason about them within a shared architecture. That consistency is especially valuable when the same capability must serve a customer interface, an internal application, and an automated agent.

Integrating Hybrid ID without merging authority boundaries

For a participating business application, begin with the Hybrid ID developer and SSO guides. The supported web flow uses a registered application, exact HTTPS callback URLs, approved scopes, and Authorization Code with PKCE S256. Developer Apps and the delegated developer-management API provide the published registration paths. Redirect the person to Hybrid ID instead of collecting their Hybrid password. Use the canonical issuer's discovery metadata and a maintained OIDC client to validate the response and establish your own application session.

Use issuer and subject to recognize the returning identity. Request only the claims needed by the registered integration, and distinguish consent to identity disclosure from membership in an organization or permission to use a product. The provider's pairwise subjects should not be replaced with email-based assumptions about account equivalence across unrelated applications. Keep client credentials and code exchange in the trusted server integration described by the guide.

The OIDC provider's discovery and protocol endpoints are separate from Hybrid-Chain's V2 product API catalog. Approved first-party account APIs and inbound federation routes have different purposes from outward-facing Hybrid ID SSO. After sign-in, confirm the authority required by each product operation; an OIDC session is not a universal wallet, vault, or workload credential. Use the current application-registration and agent contracts to establish their exact permissions and availability.

A useful integration exercise checks a returning user's own workspace, a user without the required organizational membership, and an unavailable or withdrawn product permission. The application should explain each result without leaking another organization's data or silently giving an agent additional powers. This tests the connection between identity and business access while preserving each service's independent responsibility.

Integrate each identity capability through its own contract

IntegrationUseful continuityBoundary to preserve
SSO application registrationA registered application can establish a validated returning identity.App-management credentials configure the application; they do not impersonate a user.
Guest linkingA visitor can keep application progress after approved sign-in.Require the supported possession and consent checks; never merge by email.
Application profile dataSmall preferences and context can follow the returning app/user pair.Explicit storage permission, per-app isolation, and version conflicts remain relevant.
Agent workload and delegationA runtime can be associated with an owner and a bounded task.Knowledge grants, application admission, and financial execution authority remain separate.

Use the public Markdown guides and machine-readable schemas linked by Hybrid ID to plan the integration. Keep first-party owner-portal operations, third-party developer APIs, and domain V2 contracts in their documented authentication context. A documented internal management route is not an invitation to automate a person's password or collect their authenticator secret.

Cortex and Banking add application-specific records around the shared infrastructure. Preserve those applications' admission, confidentiality, and source-of-truth rules rather than attempting to reproduce their authority in a generic connector. Agree the exact operation, effective principal, audience, retry behavior, and acceptance evidence before wiring a cross-application handoff.