Files
certctl/internal/service/auth/actor_role_service.go
T
shankar0123 3ef45e2ad4 auth-bundle-1 Phase 6-7-8: bootstrap path + scope-down CLI + auditor-role split
# Phase 6 — day-0 admin bootstrap

* internal/auth/bootstrap/ (new package): Strategy interface +
  EnvTokenStrategy with constant-time compare, one-shot consumption
  via sync.Mutex, optional admin-existence probe. Bundle 2's OIDC-
  first-admin will plug in alongside as an alternate Strategy.
* BootstrapService.ValidateAndMint: validates the operator's
  CERTCTL_BOOTSTRAP_TOKEN, mints a 32-byte (64-hex-char) random API
  key value, persists the SHA-256 hash to api_keys, grants r-admin
  via actor_roles, AddHashed's the runtime keystore so the just-
  minted key authenticates the next request without restart, and
  records bootstrap.consume to the audit trail with category=auth.
* internal/auth/keystore.go (new): KeyStore interface +
  StaticKeyStore (immutable env-var-only path) + MutableKeyStore
  (env-var keys + DB-loaded api_keys + runtime AddHashed). The auth
  middleware now consumes a KeyStore so the bootstrap path can
  extend the lookup table at runtime.
* migrations/000031_api_keys.up/down.sql: api_keys table with
  (id, name UNIQUE, key_hash UNIQUE, tenant_id, admin, created_by,
  created_at, expires_at, last_used_at). Idempotent.
* /v1/auth/bootstrap GET (probe) + POST (mint) — auth-exempt. Both
  routes documented in api/openapi.yaml + AuthExemptRouterRoutes
  allowlist updated. The token never leaves internal/auth/bootstrap;
  the minted plaintext key flows only into the HTTP response body.
* Startup warning emitted when CERTCTL_BOOTSTRAP_TOKEN is set AND
  admin actors already exist (config drift signal).
* Tests: 4 strategy invariants (empty token born disabled, wrong
  token=ErrInvalidToken without consumption, one-shot consumption,
  admin-exists closes path), 5 service tests (happy path + actor-
  name validation + propagation of strategy errors + nil-deps
  guard + 32-byte entropy budget), 8 HTTP-handler tests (status
  201/410/401/400 mapping + token-leak hygiene scan of slog +
  audit details + Location header). Token-leak test redirects
  slog.Default to a buffer for the test scope.

# Phase 7 — API-key migration + scope-down CLI

* GET /v1/auth/keys handler + service method ListKeys backed by
  ActorRoleRepository.ListDistinctActors. Returns one row per
  (actor_id, actor_type) pair with the slice of role IDs they hold.
  Permission: auth.role.list.
* internal/cli/auth_scope_down.go: AuthListKeys, AuthScopeDown
  (interactive), AuthScopeDownNonInteractive (JSON config),
  AuthScopeDownSuggest (--suggest with optional --apply). The
  synthetic actor-demo-anon is filtered out of every interactive /
  bulk path; non-interactive flow logs and skips it explicitly.
* SuggestRoleFromAuditEvents (pure function): walks 30 days of
  audit events per actor and returns the narrowest matching role
  (admin / mcp / viewer / agent / operator) plus a one-line reason.
  Classification: any admin-shaped action wins; otherwise all-MCP
  → mcp; all-read-only → viewer; all-agent-shaped → agent;
  otherwise operator. Test table pins all six classifications.
* CLI subcommand tree extended: 'auth keys list' + 'auth keys
  scope-down [--non-interactive <cfg>] [--suggest [--apply]]'.
* CHANGELOG.md leads v2.1.0 with the SECURITY: AUDIT YOUR API KEYS
  call-out + four flow examples.

# Phase 8 — auditor role + event_category column

* migrations/000032_audit_category.up/down.sql: ALTER TABLE
  audit_events ADD COLUMN event_category TEXT NOT NULL DEFAULT
  'cert_lifecycle' + CHECK constraint (cert_lifecycle/auth/config)
  + (event_category) and (event_category, timestamp DESC) indexes
  for the auditor-filter query path. WORM trigger from migration
  000018 continues to enforce append-only at the DB layer (DDL is
  not blocked).
* domain.AuditEvent gains EventCategory string (omitempty);
  domain.EventCategoryCertLifecycle / Auth / Config constants.
* AuditService.RecordEventWithCategory sibling of RecordEvent;
  legacy callers stay on RecordEvent (defaults to cert_lifecycle).
  Auth callers (RoleService, ActorRoleService, BootstrapService)
  switched to RecordEventWithCategory(..., 'auth', ...).
* GET /v1/audit?category=<cat>: handler accepts the optional query
  param, validates against the enum (400 on invalid value),
  dispatches through ListAuditEventsByCategory. OpenAPI updated
  with the new query param + AuditEvent.event_category schema.
* Postgres AuditRepository.Create now writes event_category;
  AuditRepository.List filters on it; AuditFilter.EventCategory
  gates the WHERE clause.
* Tests: 5 audit-category-filter HTTP tests (dispatch routing,
  back-compat fallback, 400 for invalid values, all 3 enum values
  accepted, page+category combine, JSON output surfaces the
  field). 3 auditor-role invariants (auditor holds exactly
  audit.read+audit.export, no mutating perms, disjoint from
  viewer except audit.read).

# Cross-phase wiring

* HandlerRegistry.Bootstrap field added; cmd/server/main.go wires
  the bootstrap service ahead of RegisterHandlers (extracted
  assembleNamedAPIKeys helper into auth_backfill.go, moved the
  keystore + bootstrap construction up alongside the auth repos).
* AuthCheckResolver / AuthActorRoleService extended with ListKeys
  to satisfy the Phase 7 surface; existing fakes updated.
* fakeAudit + mockAuditService stubs in tests gain
  RecordEventWithCategory + ListAuditEventsByCategory; existing
  tests untouched.

# Verifications

* gofmt -l: clean across every modified file.
* go vet ./...: clean.
* staticcheck across internal/auth + handler + router + cli +
  service + repository + cmd + domain: clean.
* go test -short -count=1: green across every Bundle-1-touched
  package — internal/auth (incl. bootstrap), internal/api/handler,
  internal/api/router, internal/cli, internal/service/auth,
  internal/service, internal/domain/auth, internal/repository/postgres,
  cmd/server, cmd/cli, plus internal/scheduler, internal/api/middleware,
  cmd/agent, internal/mcp.
2026-05-09 20:15:43 +00:00

178 lines
6.5 KiB
Go

package auth
import (
"context"
"fmt"
"github.com/certctl-io/certctl/internal/domain"
authdomain "github.com/certctl-io/certctl/internal/domain/auth"
"github.com/certctl-io/certctl/internal/repository"
)
// ActorRoleService grants / revokes roles to actors and exposes the
// effective-permissions query the Phase 3 middleware uses on the hot
// path.
type ActorRoleService struct {
repo repository.ActorRoleRepository
roleRepo repository.RoleRepository
authorizer *Authorizer
audit AuditService
}
// NewActorRoleService constructs an ActorRoleService.
func NewActorRoleService(
repo repository.ActorRoleRepository,
roleRepo repository.RoleRepository,
authorizer *Authorizer,
audit AuditService,
) *ActorRoleService {
return &ActorRoleService{
repo: repo,
roleRepo: roleRepo,
authorizer: authorizer,
audit: audit,
}
}
// Grant assigns a role to an actor. Privilege-escalation guard: the
// caller must hold `auth.role.assign` (globally). System callers
// bypass. Reserved actor `actor-demo-anon` is rejected.
func (s *ActorRoleService) Grant(ctx context.Context, caller *Caller, ar *authdomain.ActorRole) error {
if caller == nil {
return ErrUnauthenticated
}
if !caller.IsSystem {
ok, err := s.authorizer.HoldsAnyOf(ctx, caller.ActorID, authdomain.ActorTypeValue(caller.ActorType), s.tenantOf(caller), "auth.role.assign")
if err != nil {
return err
}
if !ok {
return fmt.Errorf("%w: auth.role.assign required", ErrSelfRoleAssignment)
}
}
if ar.ActorID == authdomain.DemoAnonActorID {
return fmt.Errorf("%w: actor-demo-anon is reserved", repository.ErrAuthReservedActor)
}
if ar.TenantID == "" {
ar.TenantID = authdomain.DefaultTenantID
}
if err := s.repo.Grant(ctx, ar); err != nil {
return err
}
s.recordAudit(ctx, caller, "actor_role.grant", "actor_role", ar.ID, map[string]interface{}{
"actor_id": ar.ActorID,
"actor_type": string(ar.ActorType),
"role_id": ar.RoleID,
})
return nil
}
// Revoke removes a previously-granted role from an actor. Same
// privilege guard as Grant: caller needs `auth.role.assign` to mutate
// role membership. Reserved actor `actor-demo-anon` is rejected so the
// demo path stays alive even after a misclick.
func (s *ActorRoleService) Revoke(ctx context.Context, caller *Caller, actorID string, actorType domain.ActorType, roleID string) error {
if caller == nil {
return ErrUnauthenticated
}
if !caller.IsSystem {
ok, err := s.authorizer.HoldsAnyOf(ctx, caller.ActorID, authdomain.ActorTypeValue(caller.ActorType), s.tenantOf(caller), "auth.role.assign")
if err != nil {
return err
}
if !ok {
return fmt.Errorf("%w: auth.role.assign required", ErrSelfRoleAssignment)
}
}
if actorID == authdomain.DemoAnonActorID {
return fmt.Errorf("%w: actor-demo-anon is reserved", repository.ErrAuthReservedActor)
}
tenantID := s.tenantOf(caller)
if err := s.repo.Revoke(ctx, actorID, authdomain.ActorTypeValue(actorType), roleID, tenantID); err != nil {
return err
}
s.recordAudit(ctx, caller, "actor_role.revoke", "actor_role", roleID, map[string]interface{}{
"actor_id": actorID,
"actor_type": string(actorType),
"role_id": roleID,
})
return nil
}
// ListForActor returns the roles held by the named actor.
func (s *ActorRoleService) ListForActor(ctx context.Context, caller *Caller, actorID string, actorType domain.ActorType) ([]*authdomain.ActorRole, error) {
if caller == nil {
return nil, ErrUnauthenticated
}
if !caller.IsSystem && caller.ActorID != actorID {
ok, err := s.authorizer.HoldsAnyOf(ctx, caller.ActorID, authdomain.ActorTypeValue(caller.ActorType), s.tenantOf(caller), "auth.role.list")
if err != nil {
return nil, err
}
if !ok {
return nil, fmt.Errorf("%w: auth.role.list required to view another actor's roles", ErrForbidden)
}
}
return s.repo.ListByActor(ctx, actorID, authdomain.ActorTypeValue(actorType), s.tenantOf(caller))
}
// EffectivePermissions returns the deduplicated (permission, scope)
// pairs granted to the actor across all roles. Phase 3 middleware
// (auth.RequirePermission) calls this on every gated request via the
// Authorizer; that hot path skips RBAC self-checks. The service-level
// method here is for handler / GUI callers (the /v1/auth/me endpoint).
func (s *ActorRoleService) EffectivePermissions(ctx context.Context, caller *Caller, actorID string, actorType domain.ActorType) ([]repository.EffectivePermission, error) {
if caller == nil {
return nil, ErrUnauthenticated
}
if !caller.IsSystem && caller.ActorID != actorID {
ok, err := s.authorizer.HoldsAnyOf(ctx, caller.ActorID, authdomain.ActorTypeValue(caller.ActorType), s.tenantOf(caller), "auth.role.list")
if err != nil {
return nil, err
}
if !ok {
return nil, fmt.Errorf("%w: auth.role.list required to view another actor's permissions", ErrForbidden)
}
}
return s.repo.EffectivePermissions(ctx, actorID, authdomain.ActorTypeValue(actorType), s.tenantOf(caller))
}
// ListKeys (Bundle 1 Phase 7) returns every actor in the tenant that
// holds at least one role grant. Permission `auth.role.list` is
// required (or the caller must be system). The CLI's `auth keys list`
// + scope-down helper consume this to enumerate the operator-key
// population without a separate /v1/auth/keys-by-name surface.
func (s *ActorRoleService) ListKeys(ctx context.Context, caller *Caller) ([]repository.ActorWithRoles, error) {
if caller == nil {
return nil, ErrUnauthenticated
}
if !caller.IsSystem {
ok, err := s.authorizer.HoldsAnyOf(ctx, caller.ActorID, authdomain.ActorTypeValue(caller.ActorType), s.tenantOf(caller), "auth.role.list")
if err != nil {
return nil, err
}
if !ok {
return nil, fmt.Errorf("%w: auth.role.list required to list keys", ErrForbidden)
}
}
return s.repo.ListDistinctActors(ctx, s.tenantOf(caller))
}
func (s *ActorRoleService) tenantOf(caller *Caller) string {
if caller != nil && caller.TenantID != "" {
return caller.TenantID
}
return authdomain.DefaultTenantID
}
func (s *ActorRoleService) recordAudit(ctx context.Context, caller *Caller, action, resourceType, resourceID string, details map[string]interface{}) {
if s.audit == nil || caller == nil {
return
}
// Bundle 1 Phase 8: every actor-role grant/revoke is an
// authentication / authorization event. The auditor role queries
// /v1/audit?category=auth to surface this slice without
// also pulling in cert.* events.
_ = s.audit.RecordEventWithCategory(ctx, caller.ActorID, caller.ActorType, action, domain.EventCategoryAuth, resourceType, resourceID, details)
}