Files
certctl/internal/auth/keystore.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

158 lines
5.4 KiB
Go

package auth
import (
"crypto/subtle"
"sync"
)
// KeyStore is the lookup contract NewAuthWithKeyStore consults to
// resolve a Bearer token (already SHA-256 hashed by the middleware) to
// a NamedAPIKey identity. The interface exists so the same auth
// middleware can serve both the env-var-keys-only path (immutable
// in-memory hash table built at startup) and the bootstrap-extended
// path (env-var keys plus runtime-minted admin keys persisted in
// `api_keys`). Bundle 2 will plug in an OIDC-session lookup behind the
// same interface.
//
// LookupByHash MUST be safe for concurrent reads. Implementations that
// support runtime additions wrap their backing slice/map in a
// sync.RWMutex (see MutableKeyStore) so the request path remains lock-
// free in the steady state.
type KeyStore interface {
// LookupByHash returns the NamedAPIKey whose SHA-256 hash matches
// the supplied hex-encoded hash. The matched bool is false when no
// entry matches; callers MUST treat false as "wrong key" (HTTP
// 401) and never as "fall through to a default identity".
//
// The supplied hash is the output of HashAPIKey(token) — already a
// 64-char lowercase hex string. Implementations compare it against
// stored hashes via crypto/subtle.ConstantTimeCompare so a
// timing-attacking caller can't byte-by-byte recover a key.
LookupByHash(hash string) (NamedAPIKey, bool)
}
// StaticKeyStore is the immutable Bundle-0 behaviour: the entries are
// fixed at construction and the lookup is a constant-time scan. Used
// by deployments that haven't enabled the Bundle-1 bootstrap flow and
// by tests that don't need runtime additions.
type StaticKeyStore struct {
entries []entry
}
type entry struct {
hash string // SHA-256 hex
name string
admin bool
}
// NewStaticKeyStore builds an immutable KeyStore from a slice of
// NamedAPIKey values. Each key is hashed once at construction. The
// returned store is safe for concurrent reads with no locking; mutation
// is not supported.
func NewStaticKeyStore(keys []NamedAPIKey) *StaticKeyStore {
out := &StaticKeyStore{
entries: make([]entry, 0, len(keys)),
}
for _, nk := range keys {
out.entries = append(out.entries, entry{
hash: HashAPIKey(nk.Key),
name: nk.Name,
admin: nk.Admin,
})
}
return out
}
// LookupByHash implements KeyStore.
func (s *StaticKeyStore) LookupByHash(hash string) (NamedAPIKey, bool) {
for i := range s.entries {
if subtle.ConstantTimeCompare([]byte(hash), []byte(s.entries[i].hash)) == 1 {
e := s.entries[i]
return NamedAPIKey{Name: e.name, Admin: e.admin}, true
}
}
return NamedAPIKey{}, false
}
// Len reports how many entries the store holds. Test/debug helper; the
// request path uses LookupByHash which is the load-bearing contract.
func (s *StaticKeyStore) Len() int { return len(s.entries) }
// MutableKeyStore is the Bundle-1 Phase 6 KeyStore that supports
// runtime additions. The Bundle 1 bootstrap flow inserts a new row
// into `api_keys`, then calls Add(...) so the just-minted key
// authenticates the very next request without a server restart. The
// backing store loads the same `api_keys` rows on startup so DB-
// persisted keys survive process restart.
//
// Concurrency: a sync.RWMutex guards a slice of entries. Reads
// (LookupByHash) take the read lock; Add takes the write lock. The
// in-memory slice mirrors the env-var named-key entries plus every
// `api_keys` row loaded at boot plus every Add that fires after
// startup.
type MutableKeyStore struct {
mu sync.RWMutex
entries []entry
}
// NewMutableKeyStore seeds a MutableKeyStore with the provided keys.
// Pass the env-var named keys here at boot; Add additional keys
// (loaded from `api_keys` or minted by bootstrap) after construction.
func NewMutableKeyStore(seed []NamedAPIKey) *MutableKeyStore {
out := &MutableKeyStore{
entries: make([]entry, 0, len(seed)),
}
for _, nk := range seed {
out.entries = append(out.entries, entry{
hash: HashAPIKey(nk.Key),
name: nk.Name,
admin: nk.Admin,
})
}
return out
}
// LookupByHash implements KeyStore.
func (s *MutableKeyStore) LookupByHash(hash string) (NamedAPIKey, bool) {
s.mu.RLock()
defer s.mu.RUnlock()
for i := range s.entries {
if subtle.ConstantTimeCompare([]byte(hash), []byte(s.entries[i].hash)) == 1 {
e := s.entries[i]
return NamedAPIKey{Name: e.name, Admin: e.admin}, true
}
}
return NamedAPIKey{}, false
}
// Add registers a new key with the store. The plaintext key is hashed
// once and stored alongside the name + admin flag. Idempotent on
// duplicate hashes (an existing entry for the same hash is replaced
// in-place so re-running the bootstrap loader on startup is safe).
func (s *MutableKeyStore) Add(key NamedAPIKey) {
s.AddHashed(key.Name, HashAPIKey(key.Key), key.Admin)
}
// AddHashed registers a key whose SHA-256 hash is already computed.
// Used by the api_keys boot loader (the DB stores the hash, not the
// plaintext, so the loader has no plaintext to re-hash).
func (s *MutableKeyStore) AddHashed(name, hashHex string, admin bool) {
s.mu.Lock()
defer s.mu.Unlock()
for i := range s.entries {
if s.entries[i].hash == hashHex {
s.entries[i].name = name
s.entries[i].admin = admin
return
}
}
s.entries = append(s.entries, entry{hash: hashHex, name: name, admin: admin})
}
// Len reports the current entry count. Test helper.
func (s *MutableKeyStore) Len() int {
s.mu.RLock()
defer s.mu.RUnlock()
return len(s.entries)
}