Skip to content

How SigID Works

SigID separates global identity from tenant-local application access. Users can reuse one account while tenants retain control over membership, roles, policies, applications, organizations, agents, wallets, billing, and audit logs.

Core objects

Object Meaning
User A person with a reusable SigID account
Tenant Application or organization boundary using SigID
Application OAuth client configured by a tenant
Organization Shared account or business unit inside a tenant
Agent Software principal that can authenticate and receive tokens
Scope Permission requested by an application, API, or agent
Role Named access bundle assigned by a tenant or organization
Policy Conditional authorization rule
Consent User approval for data sharing or delegated access
Delegation Reduced authority for an agent or service to act for another subject
Webhook Signed event delivery to a tenant receiver
Audit log Tenant-visible record of high-impact events

Global identity, tenant-local access

Users have one SigID account, but each tenant receives a tenant-local view. This protects user privacy and prevents applications from correlating users across tenants without explicit sharing.

Global SigID account
  -> tenant A subject: sub_a
  -> tenant B subject: sub_b
  -> tenant C subject: sub_c

Applications should store the tenant-local subject. Email, phone, social account, and wallet data can change or may not be shared with every tenant.

Hosted auth

Applications redirect users to SigID for authentication. SigID handles sign-in, MFA, consent, and token issuance, then redirects back to the application.

Hosted auth provides:

  • consistent sign-in methods
  • passkey and MFA enforcement
  • SSO routing
  • consent capture
  • tenant branding
  • session and recovery flows
  • OAuth and OIDC protocol handling

Tenant control plane

Tenant administrators use the dashboard and APIs to manage:

  • applications
  • redirect URIs and origins
  • users and invitations
  • organizations and SSO
  • roles, scopes, and policies
  • agents and identity anchors
  • managed wallets
  • credential vault grants
  • webhooks
  • billing and usage
  • audit logs

Authorization model

SigID combines protocol-level scopes with tenant-specific authorization. For the full authorization sequence, see Applications And OAuth.

This lets your application enforce simple scope checks for common API calls while still supporting stricter policy for sensitive actions.

Delegation

Delegation lets an agent or service act with reduced authority. SigID narrows access using user consent, tenant policy, requested scopes, subject type, and time limits.

Delegated access should answer four questions:

Question Example
Who is acting? Agent, service, or application client
For whom? User, organization, tenant, or agent itself
What can it do? Scopes, tool capabilities, or wallet actions
For how long? Token lifetime, delegation expiry, or task boundary

Events and auditability

Tenant administrators can use audit logs and webhooks to review sensitive actions such as:

  • user access changes
  • SSO updates
  • webhook changes
  • agent activity
  • key and anchor changes
  • vault access
  • wallet signing
  • policy changes
  • billing and commerce events

Audit logs are for investigation and compliance. Webhooks are for integrating SigID changes into your own systems.

Mental model

SigID is not only "login." It is a tenant-scoped identity control plane:

Need SigID surface
Let users sign in Hosted auth and OAuth
Protect APIs Access tokens, scopes, policies
Manage customers Organizations and SSO
Manage people Tenant users and roles
Manage automation Agents, delegations, vault grants
Track changes Audit logs and webhooks
Reduce user correlation Pairwise tenant-local subjects