Configuration
Configure ThreatFade's identity boundary, tenancy, persistence and allowed origins.
Configuration
ThreatFade separates development convenience from the production security boundary. Local development is intentionally permissive; production authentication is fail-closed and requires the configured identity boundary.
Production OIDC
The engine documents these settings:
THREATFADE_OIDC_ISSUER
THREATFADE_OIDC_AUDIENCE
THREATFADE_OIDC_JWKS_URL # optional; issuer discovery is used when omitted
Production tokens must contain sub, exp, iat and a tenant claim using the repository's documented tenant naming convention.
Register exact redirect URIs, enforce TLS, protect tokens and use PKCE where applicable. Do not treat a local-development configuration as a production identity configuration.
Allowed origins
A production deployment should explicitly restrict browser origins. The repository's reference deployment uses THREATFADE_ALLOWED_ORIGINS for this boundary.
export THREATFADE_ALLOWED_ORIGINS='https://console.example.com'
Tenancy and roles
The documented RBAC roles are:
vieweranalystapi_onlyadmintenant_admin
Detection records and audit events are tenant-scoped. Tenant-admin and platform-admin responsibilities are separated.
Persistence
- SQLite is documented for development.
- PostgreSQL is the production persistence path through SQLAlchemy.
- Production deployments should use durable storage and a secret-management system.
Operational targets
The readiness endpoint exposes configured SLO targets. Targets are targets, not measured guarantees. Production teams should establish evidence for availability, latency, throughput, error rate, alert volume and resource utilization before publishing measured SLOs.