up
Some checks failed
Build Test Deploy / docs (push) Has been cancelled
Build Test Deploy / deploy (push) Has been cancelled
Build Test Deploy / build-test (push) Has been cancelled
Build Test Deploy / authority-container (push) Has been cancelled
Docs CI / lint-and-preview (push) Has been cancelled

This commit is contained in:
2025-10-12 20:37:18 +03:00
parent b97fc7685a
commit 607e72e2a1
306 changed files with 21409 additions and 4449 deletions

View File

@@ -1,6 +1,6 @@
# Authority Plug-in Developer Guide
> **Status:** Ready for Docs/DOC4 editorial review as of 2025-10-10. Content aligns with PLG6 acceptance criteria and references stable Authority primitives.
> **Status:** Updated 2025-10-11 (AUTHPLUG-DOCS-01-001) with lifecycle + limiter diagrams and refreshed rate-limit guidance aligned to PLG6 acceptance criteria.
## 1. Overview
Authority plug-ins extend the **StellaOps Authority** service with custom identity providers, credential stores, and client-management logic. Unlike Feedser plug-ins (which ingest or export advisories), Authority plug-ins participate directly in authentication flows:
@@ -17,6 +17,10 @@ Authority hosts follow a deterministic plug-in lifecycle. The flow below can be
3. **Registrar execution** each assembly is searched for `IAuthorityPluginRegistrar` implementations. Registrars bind options, register services, and optionally queue bootstrap tasks.
4. **Runtime** the host resolves `IIdentityProviderPlugin` instances, uses capability metadata to decide which OAuth grants to expose, and invokes health checks for readiness endpoints.
![Authority plug-in lifecycle diagram](../assets/authority/authority-plugin-lifecycle.svg)
_Source:_ `docs/assets/authority/authority-plugin-lifecycle.mmd`
**Data persistence primer:** the standard Mongo-backed plugin stores users in collections named `authority_users_<pluginName>` and lockout metadata in embedded documents. Additional plugins must document their storage layout and provide deterministic collection naming to honour the Offline Kit replication process.
## 3. Capability Metadata
@@ -100,7 +104,7 @@ Capability flags let the host reason about what your plug-in supports:
- Optional `IClientProvisioningStore` for machine-to-machine clients.
- `AuthorityIdentityProviderCapabilities` to advertise supported flows.
- Password guidance:
- Prefer Argon2 (Security Guild upcoming recommendation); Standard plug-in currently ships PBKDF2 with easy swap via `IPasswordHasher`.
- Standard plug-in hashes via `ICryptoProvider` using Argon2id by default and emits PHC-compliant strings. Successful PBKDF2 logins trigger automatic rehashes so migrations complete gradually. See `docs/security/password-hashing.md` for tuning advice.
- Enforce password policies before hashing to avoid storing weak credentials.
- Health checks should probe backing stores (e.g., Mongo `ping`) and return `AuthorityPluginHealthResult` so `/ready` can surface issues.
- When supporting additional factors (e.g., TOTP), implement `SupportsMfa` and document the enrolment flow for resource servers.
@@ -111,14 +115,64 @@ Capability flags let the host reason about what your plug-in supports:
- Never store raw secrets in git: allow operators to supply them via `.local.yaml`, environment variables, or injected secret files. Document which keys are mandatory.
- Validate configuration as soon as the registrar runs; use explicit error messages to guide operators. The Standard plug-in now enforces complete bootstrap credentials (username + password) and positive lockout windows via `StandardPluginOptions.Validate`.
- Cross-reference bootstrap workflows with `docs/ops/authority_bootstrap.md` (to be published alongside CORE6) so operators can reuse the same payload formats for manual provisioning.
- `passwordHashing` inherits defaults from `authority.security.passwordHashing`. Override only when hardware constraints differ per plug-in:
```yaml
passwordHashing:
algorithm: Argon2id
memorySizeInKib: 19456
iterations: 2
parallelism: 1
```
Invalid values (≤0) fail fast during startup, and legacy PBKDF2 hashes rehash automatically once the new algorithm succeeds.
## 8. Logging, Metrics, and Diagnostics
### 7.1 Token Persistence Contract
- The host automatically persists every issued principal (access, refresh, device, authorization code) in `authority_tokens`. Plug-in code **must not** bypass this store; use the provided `IAuthorityTokenStore` helpers when implementing custom flows.
- When a plug-in disables a subject or client outside the standard handlers, call `IAuthorityTokenStore.UpdateStatusAsync(...)` for each affected token so revocation bundles stay consistent.
- Supply machine-friendly `revokedReason` codes (`compromised`, `rotation`, `policy`, `lifecycle`, etc.) and optional `revokedMetadata` entries when invalidating credentials. These flow straight into `revocation-bundle.json` and should remain deterministic.
- Token scopes should be normalised (trimmed, unique, ordinal sort) before returning from plug-in verification paths. `TokenPersistenceHandlers` will keep that ordering for downstream consumers.
### 7.2 Claims & Enrichment Checklist
- Authority always sets the OpenID Connect basics: `sub`, `client_id`, `preferred_username`, optional `name`, and `role` (for password flows). Plug-ins must use `IClaimsEnricher` to append additional claims in a **deterministic** order (sort arrays, normalise casing) so resource servers can rely on stable shapes.
- Recommended enrichment keys:
- `stellaops.realm` plug-in/tenant identifier so services can scope policies.
- `stellaops.subject.type` values such as `human`, `service`, `bootstrap`.
- `groups` / `projects` sorted arrays describing operator entitlements.
- Claims visible in tokens should mirror what `/token` and `/userinfo` emit. Avoid injecting sensitive PII directly; mark values with `ClassifiedString.Personal` inside the plug-in so audit sinks can tag them appropriately.
- For client-credential flows, remember to enrich both the client principal and the validation path (`TokenValidationHandlers`) so refresh flows keep the same metadata.
### 7.3 Revocation Bundles & Reasons
- Use `IAuthorityRevocationStore` to record subject/client/token revocations when credentials are deleted or rotated. Stick to the standard categories (`token`, `subject`, `client`, `key`).
- Include a deterministic `reason` string and optional `reasonDescription` so operators understand *why* a subject was revoked when inspecting bundles offline.
- Plug-ins should populate `metadata` with stable keys (e.g., `revokedBy`, `sourcePlugin`, `ticketId`) to simplify SOC correlation. The keys must be lowercase, ASCII, and free of secrets—bundles are mirrored to air-gapped agents.
## 8. Rate Limiting & Lockout Interplay
Rate limiting and account lockouts are complementary controls. Plug-ins must surface both deterministically so operators can correlate limiter hits with credential rejections.
**Baseline quotas** (from `docs/dev/authority-rate-limit-tuning-outline.md`):
| Endpoint | Default policy | Notes |
|----------|----------------|-------|
| `/token` | 30 requests / 60s, queue 0 | Drop to 10/60s for untrusted ranges; raise only with WAF + monitoring. |
| `/authorize` | 60 requests / 60s, queue 10 | Reduce carefully; interactive UX depends on headroom. |
| `/internal/*` | Disabled by default; recommended 5/60s when enabled | Keep queue 0 for bootstrap APIs. |
**Retry metadata:** The middleware stamps `Retry-After` plus tags `authority.client_id`, `authority.remote_ip`, and `authority.endpoint`. Plug-ins should keep these tags intact when crafting responses or telemetry so dashboards remain consistent.
**Lockout counters:** Treat lockouts as **subject-scoped** decisions. When multiple instances update counters, reuse the deterministic tie-breakers documented in `src/DEDUP_CONFLICTS_RESOLUTION_ALGO.md` (freshness overrides, precedence, and stable hashes) to avoid divergent lockout states across replicas.
**Alerting hooks:** Emit structured logs/metrics when either the limiter or credential store rejects access. Suggested gauges include `aspnetcore_rate_limiting_rejections_total{limiter="authority-token"}` and any custom `auth.plugins.<pluginName>.lockouts_total` counter.
![Authority rate limit and lockout flow](../assets/authority/authority-rate-limit-flow.svg)
_Source:_ `docs/assets/authority/authority-rate-limit-flow.mmd`
## 9. Logging, Metrics, and Diagnostics
- Always log via the injected `ILogger<T>`; include `pluginName` and correlation IDs where available.
- Activity/metric names should align with `AuthorityTelemetry` constants (`service.name=stellaops-authority`).
- Expose additional diagnostics via structured logging rather than writing custom HTTP endpoints; the host will integrate these into `/health` and `/ready`.
- Emit metrics with stable names (`auth.plugins.<pluginName>.*`) when introducing custom instrumentation; coordinate with the Observability guild to reserve prefixes.
## 9. Testing & Tooling
## 10. Testing & Tooling
- Unit tests: use Mongo2Go (or similar) to exercise credential stores without hitting production infrastructure (`StandardUserCredentialStoreTests` is a template).
- Determinism: fix timestamps to UTC and sort outputs consistently; avoid random GUIDs unless stable.
- Smoke tests: launch `dotnet run --project src/StellaOps.Authority/StellaOps.Authority` with your plug-in under `PluginBinaries/Authority` and verify `/ready`.
@@ -137,13 +191,13 @@ Capability flags let the host reason about what your plug-in supports:
}
```
## 10. Packaging & Delivery
## 11. Packaging & Delivery
- Output assembly should follow `StellaOps.Authority.Plugin.<Name>.dll` so the hosts search pattern picks it up.
- Place the compiled DLL plus dependencies under `PluginBinaries/Authority` for offline deployments; include hashes/signatures in release notes (Security Guild guidance forthcoming).
- Document any external prerequisites (e.g., CA cert bundle) in your plug-in README.
- Update `etc/authority.plugins/<plugin>.yaml` samples and include deterministic SHA256 hashes for optional bootstrap payloads when distributing Offline Kit artefacts.
## 11. Checklist & Handoff
## 12. Checklist & Handoff
- ✅ Capabilities declared and validated in automated tests.
- ✅ Bootstrap workflows documented (if `bootstrap` capability used) and repeatable.
- ✅ Local smoke test + unit/integration suites green (`dotnet test`).
@@ -151,7 +205,4 @@ Capability flags let the host reason about what your plug-in supports:
- Submit the developer guide update referencing PLG6/DOC4 and tag DevEx + Docs reviewers for sign-off.
---
**Next documentation actions:**
- Add rendered architectural diagram (PlantUML/mermaid) reflecting the lifecycle above once the Docs toolkit pipeline is ready.
- Reference the LDAP RFC (`docs/rfcs/authority-plugin-ldap.md`) in the capability section once review completes.
- Sync terminology with `docs/11_AUTHORITY.md` when that chapter is published to keep glossary terms consistent.
Mermaid sources for the embedded diagrams live under `docs/assets/authority/`. Regenerate the SVG assets with your preferred renderer before committing future updates so the visuals stay in sync with the `.mmd` definitions.