refactor(iam): move handlers/iam to domains/iam (#1249)

* refactor(iam): move handlers/iam to domains/iam

Relocate IAM's implementation and its test double to the ADR-0001 home
with git mv. The package is renamed from handlers_iam to iam and its
importers use the iamdomain alias. APIs, NATS subjects, KV buckets and
behaviour are unchanged.

The cross-domain allowlist entries now target domains/iam; the two
frozen-legacy entries and the operator/cli entry, which no longer breaks
the legacy-implementation rule, are removed. The migration record states
that per-consumer IAM capabilities remain debt but were not a
prerequisite to the move.

* docs: record the IAM move against its PR

architecture/layering/allowlist_test.go

@@ -34,7 +34,7 @@ var allowlist = []allowance{
 
 	{From: "daemon", To: "domains/ec2/awsapi/instance", Debt: "Instance adapter consumers outside EC2", Reason: "daemon/eks_cp_control.go and rds_deps.go call the instance adapter; needs an EC2 capability"},
 
-	{From: "domains/acm", To: "handlers/iam", Debt: "ACM: the IAM AES-GCM helper import (ACM relocation residual)", Reason: "IAM AES-GCM helpers; needs a shared crypto boundary"},
+	{From: "domains/acm", To: "domains/iam", Debt: "ACM: the IAM AES-GCM helper import (ACM relocation residual)", Reason: "IAM AES-GCM helpers; needs a shared crypto boundary"},
 
 	{From: "domains/admission/quota", To: "domains/ec2/awsapi/eip", Debt: "domains/admission/quota (dependency, moved from handlers/quota unchanged)", Reason: "describe adapter used as a counter; needs a reader capability"},
 	{From: "domains/admission/quota", To: "domains/ec2/awsapi/instance", Debt: "domains/admission/quota (dependency, moved from handlers/quota unchanged)", Reason: "vcpu.go counts instances through the instance adapter; needs an EC2 reader capability"},
@@ -48,17 +48,17 @@ var allowlist = []allowance{
 	{From: "domains/dns", To: "domains/network/reconcile", Debt: "domains/dns", Reason: "takes the reconcile lock from network's implementation"},
 	{From: "domains/dns", To: "runtime/compute/vm", Debt: "domains/dns", Reason: "domain-to-runtime import of VM state"},
 
-	{From: "domains/ec2/awsapi/instance", To: "handlers/iam", Debt: "domains/ec2/awsapi/instance (dependency)", Reason: "IAM edge needs an IAM capability"},
+	{From: "domains/ec2/awsapi/instance", To: "domains/iam", Debt: "domains/ec2/awsapi/instance (dependency)", Reason: "IAM edge needs an IAM capability"},
 	{From: "domains/ec2/awsapi/instance", To: "runtime/compute/cache", Debt: "domains/ec2/awsapi/instance (dependency)", Reason: "node liveness and VM state; needs an EC2-owned projection"},
 	{From: "domains/ec2/awsapi/instance", To: "runtime/compute/vm", Debt: "domains/ec2/awsapi/instance (dependency)", Reason: "node liveness and VM state; needs an EC2-owned projection"},
 
 	{From: "domains/ec2/awsapi/spotinstance", To: "domains/admission/quota", Debt: "domains/ec2/awsapi/spotinstance (dependency)", Reason: "RequestSpotInstances.go uses IAMService and quota.Service"},
-	{From: "domains/ec2/awsapi/spotinstance", To: "handlers/iam", Debt: "domains/ec2/awsapi/spotinstance (dependency)", Reason: "RequestSpotInstances.go uses IAMService and quota.Service"},
+	{From: "domains/ec2/awsapi/spotinstance", To: "domains/iam", Debt: "domains/ec2/awsapi/spotinstance (dependency)", Reason: "RequestSpotInstances.go uses IAMService and quota.Service"},
 
 	{From: "domains/ec2/eip", To: "domains/network/topology", Debt: "EC2 → network implementation", Reason: "EC2 reads network implementation; clears with the EC2-owned projection"},
 
 	{From: "domains/ec2/guestmetadata", To: "domains/dns", Debt: "Guest metadata: the cross-domain and runtime imports above", Reason: "guest-metadata relocation residual; needs capabilities, projections or contracts"},
-	{From: "domains/ec2/guestmetadata", To: "handlers/iam", Debt: "Guest metadata: the cross-domain and runtime imports above", Reason: "guest-metadata relocation residual; needs capabilities, projections or contracts"},
+	{From: "domains/ec2/guestmetadata", To: "domains/iam", Debt: "Guest metadata: the cross-domain and runtime imports above", Reason: "guest-metadata relocation residual; needs capabilities, projections or contracts"},
 	{From: "domains/ec2/guestmetadata", To: "handlers/sts", Debt: "Guest metadata: the cross-domain and runtime imports above", Reason: "guest-metadata relocation residual; needs capabilities, projections or contracts"},
 	{From: "domains/ec2/guestmetadata", To: "runtime/compute/vm", Debt: "Guest metadata: the cross-domain and runtime imports above", Reason: "guest-metadata relocation residual; needs capabilities, projections or contracts"},
 
@@ -76,7 +76,7 @@ var allowlist = []allowance{
 	{From: "domains/ec2/vpc", To: "domains/network/identifiers", Debt: "EC2 → network implementation", Reason: "EC2 reads network implementation; clears with the EC2-owned projection"},
 	{From: "domains/ec2/vpc", To: "domains/network/topology", Debt: "EC2 → network implementation", Reason: "EC2 reads network implementation; clears with the EC2-owned projection"},
 
-	{From: "domains/ecr/auth", To: "handlers/iam", Debt: "Cross-domain", Reason: "IAM implementation import; needs an IAM capability"},
+	{From: "domains/ecr/auth", To: "domains/iam", Debt: "Cross-domain", Reason: "IAM implementation import; needs an IAM capability"},
 
 	{From: "domains/network/host", To: "runtime/compute/vm", Debt: "domains/network", Reason: "domain-to-runtime import"},
 	{From: "domains/network/host", To: "runtime/host/command", Debt: "domains/network", Reason: "domain-to-runtime import"},
@@ -93,7 +93,7 @@ var allowlist = []allowance{
 	{From: "domains/ochre", To: "gateway/bedrock", Debt: "gateway/bedrock", Reason: "Ochre domain core still lives in gateway/bedrock; needs the recorded split"},
 	{From: "domains/ochre", To: "runtime/compute/gpu", Debt: "Domain → runtime", Reason: "GPU runtime state read directly; needs a domain-owned projection"},
 
-	{From: "domains/ochre/vector", To: "handlers/iam", Debt: "Cross-domain", Reason: "consumes another domain's implementation rather than a capability"},
+	{From: "domains/ochre/vector", To: "domains/iam", Debt: "Cross-domain", Reason: "consumes another domain's implementation rather than a capability"},
 
 	{From: "gateway", To: "domains/ec2/awsapi", Debt: "gateway/ec2.go (registration)", Reason: "EC2 dispatch map stays in gateway until the ingress registration seam"},
 	{From: "gateway", To: "domains/ec2/awsapi/account", Debt: "gateway/ec2.go (registration)", Reason: "EC2 dispatch map stays in gateway until the ingress registration seam"},
@@ -116,17 +116,17 @@ var allowlist = []allowance{
 	{From: "gateway", To: "domains/ec2/awsapi/vpc", Debt: "gateway/ec2.go (registration)", Reason: "EC2 dispatch map stays in gateway until the ingress registration seam"},
 	{From: "gateway", To: "domains/ec2/awsapi/zone", Debt: "gateway/ec2.go (registration)", Reason: "EC2 dispatch map stays in gateway until the ingress registration seam"},
 
-	{From: "gateway/bedrock", To: "handlers/iam", Debt: "IAM implementation consumers in STS, Ochre and ELBv2", Reason: "secret encrypt/decrypt helpers; needs a shared crypto boundary"},
+	{From: "gateway/bedrock", To: "domains/iam", Debt: "IAM implementation consumers in STS, Ochre and ELBv2", Reason: "secret encrypt/decrypt helpers; needs a shared crypto boundary"},
 
 	{From: "gateway/rds", To: "domains/ec2/awsapi/instance", Debt: "Instance adapter consumers outside EC2", Reason: "enginecatalog.go calls the instance adapter; needs an EC2 capability"},
 
-	{From: "gateway/sts", To: "handlers/iam", Debt: "IAM implementation consumers in STS, Ochre and ELBv2", Reason: "GetCallerIdentity takes IAMService; needs an IAM capability"},
+	{From: "gateway/sts", To: "domains/iam", Debt: "IAM implementation consumers in STS, Ochre and ELBv2", Reason: "GetCallerIdentity takes IAMService; needs an IAM capability"},
 
 	{From: "handlers/ecs", To: "domains/ec2/ebs/metadata", Debt: "close-out: handlers/ecs (ADR-0004 blocked-dependency list; inventory-ecs boundary imports)", Reason: "legacy ECS calls another domain's implementation as an internal API"},
 	{From: "handlers/ecs", To: "domains/ec2/eip", Debt: "close-out: handlers/ecs (ADR-0004 blocked-dependency list; inventory-ecs boundary imports)", Reason: "legacy ECS calls another domain's implementation as an internal API"},
 	{From: "handlers/ecs", To: "domains/ec2/instancetypes", Debt: "close-out: handlers/ecs (ADR-0004 blocked-dependency list; inventory-ecs boundary imports)", Reason: "legacy ECS calls another domain's implementation as an internal API"},
 	{From: "handlers/ecs", To: "handlers/elbv2", Debt: "close-out: handlers/ecs (ADR-0004 blocked-dependency list; inventory-ecs boundary imports)", Reason: "legacy ECS calls another domain's implementation as an internal API"},
-	{From: "handlers/ecs", To: "handlers/iam", Debt: "close-out: handlers/ecs (ADR-0004 blocked-dependency list; inventory-ecs boundary imports)", Reason: "legacy ECS calls another domain's implementation as an internal API"},
+	{From: "handlers/ecs", To: "domains/iam", Debt: "close-out: handlers/ecs (ADR-0004 blocked-dependency list; inventory-ecs boundary imports)", Reason: "legacy ECS calls another domain's implementation as an internal API"},
 
 	{From: "handlers/eks", To: "domains/dns", Debt: "close-out: handlers/eks (ADR-0004 blocked-dependency list; inventory-eks boundary imports)", Reason: "legacy EKS calls another domain's or runtime implementation directly"},
 	{From: "handlers/eks", To: "domains/ec2/ebs/metadata", Debt: "close-out: handlers/eks (ADR-0004 blocked-dependency list; inventory-eks boundary imports)", Reason: "legacy EKS calls another domain's or runtime implementation directly"},
@@ -134,7 +134,7 @@ var allowlist = []allowance{
 	{From: "handlers/eks", To: "domains/ec2/placementgroup", Debt: "close-out: handlers/eks (ADR-0004 blocked-dependency list; inventory-eks boundary imports)", Reason: "legacy EKS calls another domain's or runtime implementation directly"},
 	{From: "handlers/eks", To: "domains/ec2/systeminstance", Debt: "close-out: handlers/eks (ADR-0004 blocked-dependency list; inventory-eks boundary imports)", Reason: "legacy EKS calls another domain's or runtime implementation directly"},
 	{From: "handlers/eks", To: "domains/network/systemvpc", Debt: "close-out: handlers/eks (ADR-0004 blocked-dependency list; inventory-eks boundary imports)", Reason: "legacy EKS calls another domain's or runtime implementation directly"},
-	{From: "handlers/eks", To: "handlers/iam", Debt: "close-out: handlers/eks (ADR-0004 blocked-dependency list; inventory-eks boundary imports)", Reason: "legacy EKS calls another domain's or runtime implementation directly"},
+	{From: "handlers/eks", To: "domains/iam", Debt: "close-out: handlers/eks (ADR-0004 blocked-dependency list; inventory-eks boundary imports)", Reason: "legacy EKS calls another domain's or runtime implementation directly"},
 	{From: "handlers/eks", To: "runtime/compute/vm", Debt: "close-out: handlers/eks (ADR-0004 blocked-dependency list; inventory-eks boundary imports)", Reason: "legacy EKS calls another domain's or runtime implementation directly"},
 
 	{From: "handlers/elbv2", To: "domains/acm", Debt: "handlers/elbv2 (cross-domain)", Reason: "legacy ELBv2 cross-domain import; clears when ELBv2 is extracted behind consumer-owned capabilities"},
@@ -142,7 +142,7 @@ var allowlist = []allowance{
 	{From: "handlers/elbv2", To: "domains/ec2/systeminstance", Debt: "handlers/elbv2 (cross-domain)", Reason: "legacy ELBv2 cross-domain import; clears when ELBv2 is extracted behind consumer-owned capabilities"},
 	{From: "handlers/elbv2", To: "domains/ec2/vpc", Debt: "handlers/elbv2 (cross-domain)", Reason: "legacy ELBv2 cross-domain import; clears when ELBv2 is extracted behind consumer-owned capabilities"},
 	{From: "handlers/elbv2", To: "domains/network/topology", Debt: "handlers/elbv2 (cross-domain)", Reason: "legacy ELBv2 cross-domain import; clears when ELBv2 is extracted behind consumer-owned capabilities"},
-	{From: "handlers/elbv2", To: "handlers/iam", Debt: "IAM implementation consumers in STS, Ochre and ELBv2", Reason: "system-instance role ensurer; needs an IAM capability"},
+	{From: "handlers/elbv2", To: "domains/iam", Debt: "IAM implementation consumers in STS, Ochre and ELBv2", Reason: "system-instance role ensurer; needs an IAM capability"},
 	{From: "handlers/elbv2", To: "lbagent", Debt: "handlers/elbv2/{agent_types,haproxy,health_checker,nginx}.go import lbagent", Reason: "health-report and cert/PID-path vocabulary; needs a guest/controller contract"},
 
 	{From: "handlers/rds", To: "domains/dns", Debt: "close-out: handlers/rds (ADR-0004 blocked-dependency list; inventory-rds boundary imports)", Reason: "legacy RDS calls another domain's or runtime implementation directly"},
@@ -150,15 +150,14 @@ var allowlist = []allowance{
 	{From: "handlers/rds", To: "domains/ec2/instancetypes", Debt: "close-out: handlers/rds (ADR-0004 blocked-dependency list; inventory-rds boundary imports)", Reason: "legacy RDS calls another domain's or runtime implementation directly"},
 	{From: "handlers/rds", To: "domains/ec2/systeminstance", Debt: "close-out: handlers/rds (ADR-0004 blocked-dependency list; inventory-rds boundary imports)", Reason: "legacy RDS calls another domain's or runtime implementation directly"},
 	{From: "handlers/rds", To: "domains/network/systemvpc", Debt: "close-out: handlers/rds (ADR-0004 blocked-dependency list; inventory-rds boundary imports)", Reason: "legacy RDS calls another domain's or runtime implementation directly"},
-	{From: "handlers/rds", To: "handlers/iam", Debt: "close-out: handlers/rds (ADR-0004 blocked-dependency list; inventory-rds boundary imports)", Reason: "legacy RDS calls another domain's or runtime implementation directly"},
+	{From: "handlers/rds", To: "domains/iam", Debt: "close-out: handlers/rds (ADR-0004 blocked-dependency list; inventory-rds boundary imports)", Reason: "legacy RDS calls another domain's or runtime implementation directly"},
 	{From: "handlers/rds", To: "runtime/compute/vm", Debt: "close-out: handlers/rds (ADR-0004 blocked-dependency list; inventory-rds boundary imports)", Reason: "legacy RDS calls another domain's or runtime implementation directly"},
 
-	{From: "handlers/sts", To: "handlers/iam", Debt: "IAM implementation consumers in STS, Ochre and ELBv2", Reason: "STS reads IAM service, trust-policy and OIDC-provider state; needs an IAM capability"},
+	{From: "handlers/sts", To: "domains/iam", Debt: "IAM implementation consumers in STS, Ochre and ELBv2", Reason: "STS reads IAM service, trust-policy and OIDC-provider state; needs an IAM capability"},
 
 	{From: "operator/cli", To: "accountteardown", Debt: "operator/cli (dependency)", Reason: "command tree imports implementations; needs public capabilities"},
 	{From: "operator/cli", To: "daemon", Debt: "operator/cli (dependency)", Reason: "command tree imports implementations; needs public capabilities"},
 	{From: "operator/cli", To: "gateway", Debt: "operator/cli (dependency)", Reason: "command tree imports implementations; needs public capabilities"},
 	{From: "operator/cli", To: "gateway/bedrock", Debt: "operator/cli (dependency)", Reason: "command tree imports implementations; needs public capabilities"},
 	{From: "operator/cli", To: "handlers/eks", Debt: "operator/cli (dependency)", Reason: "command tree imports implementations; needs public capabilities"},
-	{From: "operator/cli", To: "handlers/iam", Debt: "operator/cli (dependency)", Reason: "command tree imports implementations; needs public capabilities"},
 }

architecture/layering/legacy_test.go

@@ -24,8 +24,6 @@ var frozenLegacyPackages = []string{
 	"handlers/ecs",
 	"handlers/eks",
 	"handlers/elbv2",
-	"handlers/iam",
-	"handlers/iam/mock",
 	"handlers/rds",
 	"handlers/sts",
 	"lbagent",

contracts/admission/v1/README.md

@@ -0,0 +1,42 @@
+# Admission contract, version 1
+
+This package owns the cross-process contract for the admission domain's
+quota override. It is not an in-process service interface or the owner of
+quota state.
+
+## Seal account quota override
+
+`SealAccountQuotaOverrideSubject` (`admission.v1.SealAccountQuotaOverride`) is
+a request/reply route on which a gateway, in the `awsgw-quota-seal` queue
+group, irreversibly seals one account's quota override record.
+`domains/admission/quota` owns it: `awsgw` serves it.
+`accountteardown` is the consumer: it calls this subject to fence every
+override writer at the quota owner before the account record is removed.
+
+The request names the account in its body (`account_id`) and may carry an
+opaque `correlation_ref` that is echoed for logs.
+`awsgw` ignores `X-Account-ID` for this subject: the account sealed is the one
+named in the body.
+A seal is refused with `IncorrectState` unless IAM already holds the account
+as `TERMINATING`; a precondition read failure, or no termination checker
+wired, answers `InternalError`.
+Both refusals write nothing and never open the override bucket.
+
+A reply is sent only on success: `disposition` is always `sealed` and
+`revision` is the stream leader's accepted write for that seal.
+A failure is the standard awserrors error payload, not this reply shape.
+Sealing is idempotent: a seal on an already-sealed record re-writes the same
+value and returns a fresh, higher revision.
+Any current gateway may answer; the leader-accepted revision is the evidence
+whichever one does, and the override bucket is opened at seal time even on a
+gateway whose quotas were disabled at startup.
+
+### Compatibility boundary
+
+The subject name carries the contract major; `account_quota_seal_test.go`
+preserves the version-1 JSON shape.
+Producer and consumer are in the coordinated Spinifex release, so the window
+is N and N-1.
+A gateway that predates the subject has no subscription, so the request gets
+no responder.
+An incompatible change to the subject or fields needs a new major.

contracts/admission/v1/account_quota_seal.go

@@ -0,0 +1,31 @@
+// Package admissionv1 holds the admission domain's cross-process contract:
+// the subjects, request/reply shapes and wire vocabulary the quota owner and
+// its consumers share.
+package admissionv1
+
+// SealAccountQuotaOverrideSubject is the request/reply subject on which a
+// gateway seals one account's quota override record. Sealing is irreversible
+// and refused unless IAM already holds the account as TERMINATING.
+const SealAccountQuotaOverrideSubject = "admission.v1.SealAccountQuotaOverride"
+
+// AccountQuotaOverrideDispositionSealed is the only disposition a successful
+// reply carries: a reply is sent only on success, so there is no other value.
+const AccountQuotaOverrideDispositionSealed = "sealed"
+
+// SealAccountQuotaOverrideRequest asks for one account's quota override to be
+// sealed. CorrelationRef is an opaque caller reference echoed for logs; it
+// grants nothing and is never a fence or authority to write.
+type SealAccountQuotaOverrideRequest struct {
+	AccountID      string `json:"account_id"`
+	CorrelationRef string `json:"correlation_ref,omitempty"`
+}
+
+// SealAccountQuotaOverrideReply confirms accountID's quota override record
+// was sealed at Revision, the stream leader's accepted write for that seal.
+// A failure is the standard awserrors error payload, not this type.
+type SealAccountQuotaOverrideReply struct {
+	AccountID      string `json:"account_id"`
+	CorrelationRef string `json:"correlation_ref,omitempty"`
+	Disposition    string `json:"disposition"`
+	Revision       uint64 `json:"revision"`
+}

contracts/admission/v1/account_quota_seal_test.go

@@ -0,0 +1,43 @@
+package admissionv1
+
+import (
+	"encoding/json"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+)
+
+func TestSealAccountQuotaOverrideSubject(t *testing.T) {
+	require.Equal(t, "admission.v1.SealAccountQuotaOverride", SealAccountQuotaOverrideSubject)
+}
+
+func TestSealAccountQuotaOverrideRequestJSONContract(t *testing.T) {
+	data, err := json.Marshal(SealAccountQuotaOverrideRequest{AccountID: "111122223333", CorrelationRef: "req-1"})
+	require.NoError(t, err)
+	require.JSONEq(t, `{"account_id":"111122223333","correlation_ref":"req-1"}`, string(data))
+
+	data, err = json.Marshal(SealAccountQuotaOverrideRequest{AccountID: "111122223333"})
+	require.NoError(t, err)
+	require.JSONEq(t, `{"account_id":"111122223333"}`, string(data))
+}
+
+func TestSealAccountQuotaOverrideReplyJSONContract(t *testing.T) {
+	reply := SealAccountQuotaOverrideReply{
+		AccountID:      "111122223333",
+		CorrelationRef: "req-1",
+		Disposition:    AccountQuotaOverrideDispositionSealed,
+		Revision:       42,
+	}
+	data, err := json.Marshal(reply)
+	require.NoError(t, err)
+	require.JSONEq(t, `{
+		"account_id": "111122223333",
+		"correlation_ref": "req-1",
+		"disposition": "sealed",
+		"revision": 42
+	}`, string(data))
+
+	data, err = json.Marshal(SealAccountQuotaOverrideReply{AccountID: "111122223333", Disposition: AccountQuotaOverrideDispositionSealed})
+	require.NoError(t, err)
+	require.JSONEq(t, `{"account_id":"111122223333","disposition":"sealed","revision":0}`, string(data))
+}

docs/PACKAGE_BOUNDARY_MIGRATION.md

@@ -340,6 +340,7 @@ while the branch is in flight; they do not describe the umbrella repository's
 | `0d830e901` | — | `spinifex/domains/tagging` (`TaggedLister`, `TaggedResource`) | The read-only list capability for the Resource Groups Tagging aggregation, declared by its consumer per ADR-0001. `ListTagged` takes the caller's account and the lower-cased `ResourceTypeFilters` set, where an entry is a full `service:type` or a bare service, and returns ARNs with read-only tag maps. The package imports no domain, handler or contract package and the layering ratchet classifies it with no allowlist row. It carries no mutation, limit or ever-tagged surface. |
 | `f3545e3ea` | `spinifex/gateway/tagging` (`listEC2`, `buildEC2Mappings`, `ec2ARN`, `elbv2Resources`, `matchesType`), copied | `spinifex/runtime/roles/awsgw` (`ec2IndexLister`, `elbv2Lister`), `spinifex/domains/tagging` (`MatchesType`) | AWSGW-side NATS client adapters implement `tagging.TaggedLister`, each with its compile-time check beside it in the composition root; no owner package carries one. The EC2 index adapter wraps `ec2.DescribeTags` and builds each ARN in the gateway's region from the stored resource type, including the unresolved-ID shapes today's code produces. The ELBv2 adapter lists load balancers and target groups admitted by the type filter, tagged or not, and reads tags in `DescribeTags` batches of 20. Both reuse existing subjects and clients, so no wire contract is added, and each call builds fresh tag maps. Type matching is exported once from `domains/tagging` so no copy survives the move. Embedded-JetStream tests cover type filters including a bare service, untagged ELBv2 resources, EC2 ARN mapping, account scoping, batch sizes and owner error propagation. |
 | `aedd0b2c4` | `spinifex/gateway/tagging` (`GetResources` parsing, tag filtering, sorting, offset token, authorization, response writing) | `spinifex/domains/tagging` (`Aggregator`, `NewAggregator`, `HasScope`, `ScopedActions`), wired by `spinifex/runtime/roles/awsgw` | The aggregation moves unchanged and consumes `[]TaggedLister`: ELBv2 is called before EC2, the non-stable ARN sort is kept, lister errors propagate with the same codes, and tag maps return as key-sorted tag lists with `[]` for untagged resources. AWSGW builds the aggregator with both adapters and sets the new `GatewayConfig.TaggingResources` field; the dispatch in `gateway/tagging.go` stays and still answers `ServerInternal` when NATS or the aggregator is not wired. The `TestCurrentBehaviour_Tagging_GetResources*` tests move to `runtime/roles/awsgw` with unchanged assertions, seeding EC2 tags through the existing EC2 tags client over NATS; pure aggregator tests move to `domains/tagging` against fake listers; the gateway keeps dispatch and error-path tests. The gateway's private copies of the listing helpers and the legacy NATS lister are deleted with their tests, `gateway/tagging` is removed, and its two allowlist rows (`gateway/tagging` → `domains/ec2/tags`, `gateway/tagging` → `handlers/elbv2`) and its frozen-legacy entry go with it. |
+| `spinifex#1249` | `spinifex/handlers/iam`, `spinifex/handlers/iam/mock` | `spinifex/domains/iam` (`iam`, imported as `iamdomain`), `spinifex/domains/iam/mock` | A mechanical relocation of IAM's implementation to its ADR-0001 home; APIs, NATS subjects, KV buckets and behaviour are unchanged and only the package name changed from `handlers_iam`. Per-consumer IAM capabilities remain dependency-rule debt but were not a prerequisite. The two frozen-legacy entries are removed, the cross-domain allowlist entries now target `domains/iam`, and the `operator/cli` → `handlers/iam` entry is removed because an operator import of a domain package does not break the legacy-implementation rule. |
 The EC2 contract row is intentionally different from the directory moves: it creates
 a compatibility boundary. `ec2.cmd.*` retains its deployed one-token NATS
 subscription and permission shape, while callers use
@@ -407,7 +408,7 @@ These residuals were recorded by individual slices and must be resolved once the
 
 - Guest metadata: the cross-domain and runtime imports above.
 - EC2 → network implementation: `domains/ec2/vpc/eni.go:18` and `eni_lease_rebind.go:11` import `domains/network/topology`, `eni.go:17` also imports `domains/network/identifiers`, `external_ipam.go:10-11` imports `domains/network/external` and `domains/network/external/dhcp`, `domains/ec2/eip/service_impl.go:17` and `lease_rebind.go:11` import `domains/network/topology`, and `domains/ec2/instance/service_impl.go:31` imports `domains/network/topology`. EC2 reads network's implementation directly rather than through a capability or projection; this clears when EC2's network-facing reads and writes go through the EC2-owned projection recorded above (step 7).
-- Cross-domain: `domains/ecr/auth/signingkey.go:24` and `domains/ochre/vector/appliance.go:15` import `handlers/iam`; this needs an IAM capability rather than the IAM implementation package. `domains/ochre/launch.go:17-18` and `domains/ochre/vpc.go:9` import `domains/ec2/systeminstance` and `domains/network/systemvpc`; Ochre consumes managed-service launch and system-VPC realisation as domain implementations rather than capabilities.
+- Cross-domain: `domains/ecr/auth/signingkey.go:24` and `domains/ochre/vector/appliance.go:15` import `domains/iam`; this needs an IAM capability rather than the IAM implementation package. `domains/ochre/launch.go:17-18` and `domains/ochre/vpc.go:9` import `domains/ec2/systeminstance` and `domains/network/systemvpc`; Ochre consumes managed-service launch and system-VPC realisation as domain implementations rather than capabilities.
 - Domain → runtime: `domains/ochre/capacity.go:9` imports `runtime/compute/gpu`; `domains/ec2/volume/service_impl.go:30` and `volume_leak_reaper.go:8` import `runtime/compute/vm`; `domains/ec2/instance/status_selection.go:9` also imports `runtime/compute/vm`. Each reads VM or GPU runtime state directly instead of through a domain-owned projection.
 - STS web-identity verification never consults cluster status: tokens verify while a cluster is DELETING until teardown sweeps the prefix, and a FAILED launch keeps its JWKS. AWS behaviour must be established before calling it a divergence. Removal condition: AWS behaviour verified, and the resolver or EKS lifecycle aligned with it.
 - The public `/keys` endpoint (root `gateway`, `oidc_discovery.go`) serves malformed stored JWKS bytes with HTTP 200. Removal condition: validate before serving, or record it as accepted.
@@ -448,8 +449,8 @@ These residuals were recorded by individual slices and must be resolved once the
 - `gateway/requestaudit.go`: the outermost request-audit middleware records AWS fields and is updated from four places in the gateway's authentication flow through the gateway's own context key. Moving it would export about seven names used only by that flow, so it stays in `gateway` until the generic ingress registration seam defines how ingress middleware receives authentication results.
 - `accountteardown` (dependency): its EC2 reapers call `domains/ec2/awsapi/*` action adapters directly, as they did under `gateway/ec2`. The ADR's `accountteardown` current-to-target map row replaces this with domain-owned cleanup capabilities before the package moves. It also holds a read-only consumer of the EC2 platform-inventory contract (`contracts/ec2/v1`, `ec2.v1.PlatformInventory`), served by `domains/ec2/accountinventory`; only the explicit dry run (`Engine.DryRun`) calls it, to log a shadow comparison, and the key-pair, placement-group and launch-template reapers still use the `awsapi` adapters. A domain-owned cleanup capability for the proven-safe subset of those three kinds (key pairs and empty placement groups no instance, spot request or live launch-template version names, and launch-template residue; never live launch templates) exists as `domains/ec2/accountcleanup` (a Go interface with injected owners and reference readers, no NATS subject); nothing calls it yet. It also holds the StageAccount settings-seal reaper, a consumer of `contracts/ec2/v1`'s `ec2.v1.SealAccountSettings` and per-node `ec2.v1.AccountSettingsSealCapability.<node>` subjects, served by `domains/ec2/account` through the daemon; the reaper reads no KV and imports no domain package, reaching both subjects only through NATS and the configured node set `NewClusterEngine` already receives.
 - `domains/admission/quota` (dependency, moved from `handlers/quota` unchanged): `live.go` imports `gateway/elbv2` and `handlers/rds` for its live ELBv2 and RDS counters; `records.go` imports `runtime/compute/vm` for the instance-record vCPU sweep; `live.go`, `vcpu.go` and `storage.go` call `domains/ec2/awsapi/{vpc,eip,volume}` describe adapters to count VPCs, subnets, addresses and volumes, as they did under `gateway/ec2`. The ADR's `handlers/quota` current-to-target map row replaces these with reader capabilities once their final locations exist.
-- `domains/ec2/awsapi/instance` (dependency): imports `handlers/iam` (`IAMService` in `DescribeInstances.go`, `RunInstances.go` and `IamInstanceProfileAssociation.go`; `InstanceProfile` in the latter two), `runtime/compute/cache` (`NodeState`, `NodeUnknown`, `NodeLive`) and `runtime/compute/vm` (`VM`), both in `status_synthesis.go`. The IAM edge needs an IAM capability; the runtime edges are domain-to-runtime and need the instance record and node liveness exposed through an EC2-owned projection.
-- `domains/ec2/awsapi/spotinstance` (dependency): imports `handlers/iam.IAMService` and `domains/admission/quota.Service` in `RequestSpotInstances.go`.
+- `domains/ec2/awsapi/instance` (dependency): imports `domains/iam` (`IAMService` in `DescribeInstances.go`, `RunInstances.go` and `IamInstanceProfileAssociation.go`; `InstanceProfile` in the latter two), `runtime/compute/cache` (`NodeState`, `NodeUnknown`, `NodeLive`) and `runtime/compute/vm` (`VM`), both in `status_synthesis.go`. The IAM edge needs an IAM capability; the runtime edges are domain-to-runtime and need the instance record and node liveness exposed through an EC2-owned projection.
+- `domains/ec2/awsapi/spotinstance` (dependency): imports `domains/iam.IAMService` and `domains/admission/quota.Service` in `RequestSpotInstances.go`.
 - Instance adapter consumers outside EC2: `gateway/rds/enginecatalog.go`, `daemon/eks_cp_control.go`, `daemon/rds_deps.go`, `domains/admission/quota`'s `vcpu.go` and `gateway/iam.go` (`CountInstanceProfileAssociations`) call `domains/ec2/awsapi/instance` directly; each needs an EC2 capability rather than the AWS action adapter.
 
 Enforcement: the owner decided the layering test is blocking now, with a ratchet. Its allowlist mirrors this debt list exactly, each edge carrying its owner and removal condition; any new unrecorded forbidden edge fails local preflight and CI, and an allowlisted edge that no longer exists must be removed from the list.
@@ -463,7 +464,7 @@ Enforcement: the owner decided the layering test is blocking now, with a ratchet
 - `agents/rds/agent` (dependency, narrowed): engine metadata now arrives through the agent-declared `EngineCatalog`, but the agent still imports `handlers/rds` for the command protocol, wire and record types (`Command*`, `Parameter`, `EngineHealth*`, `BootstrapMode*`) and API output types. This clears when those move behind a guest/controller wire contract.
 - `gateway/rds` → `handlers/rds` (`InstanceSizing`): the gateway obtains its `Sizing` through the RDS handler package's adapter rather than its own composition root. This clears when the gateway's RDS composition moves to `runtime/roles/awsgw` and builds `Sizing` there.
 - `domains/admission/quota` → `domains/ec2/instancetypes`: `reconcile.go`, `records.go` and `vcpu.go` read `DefaultVCPUs` to size instances without a recorded vCPU count. This clears when quota receives vCPU sizing through a consumer-owned capability or an EC2 projection.
-- IAM implementation consumers in STS, Ochre and ELBv2: `handlers/sts` imports `handlers/iam` across five files (`IAMService`, `SessionRevoker`, `EncryptSecret`/`DecryptSecret`, `AccessKeyStatusActive`, `ValidateTrustPolicyDocument`, `PolicyEffectAllow`/`PolicyEffectDeny`, `IAMAccountBucketName`, `OIDCProviderARN`, `OIDCProviderKey`), not only `oidc_jwks.go`; `gateway/sts/GetCallerIdentity.go` takes `IAMService`; `gateway/bedrock/credentials.go` uses `EncryptSecret`/`DecryptSecret`; and `handlers/elbv2/{lb_role,service_impl}.go` use `EnsureSystemInstanceProfile` and `SystemInstanceRoleEnsurer`. Each clears when its consumer receives a narrow IAM capability, or a shared crypto boundary for the secret helpers, rather than importing the IAM implementation.
+- IAM implementation consumers in STS, Ochre and ELBv2: `handlers/sts` imports `domains/iam` across five files (`IAMService`, `SessionRevoker`, `EncryptSecret`/`DecryptSecret`, `AccessKeyStatusActive`, `ValidateTrustPolicyDocument`, `PolicyEffectAllow`/`PolicyEffectDeny`, `IAMAccountBucketName`, `OIDCProviderARN`, `OIDCProviderKey`), not only `oidc_jwks.go`; `gateway/sts/GetCallerIdentity.go` takes `IAMService`; `gateway/bedrock/credentials.go` uses `EncryptSecret`/`DecryptSecret`; and `handlers/elbv2/{lb_role,service_impl}.go` use `EnsureSystemInstanceProfile` and `SystemInstanceRoleEnsurer`. Each clears when its consumer receives a narrow IAM capability, or a shared crypto boundary for the secret helpers, rather than importing the IAM implementation.
 - `handlers/elbv2` (cross-domain): `service_impl.go` imports `domains/acm` (`Store`, `NewStore`), `domains/dns` (`ELBChanges`, `ELBName`, `Change`, `ActionUpsert`, `ResolveBaseDomain`), `domains/ec2/vpc` (`VPCServiceImpl`) and `domains/network/topology` (`Port`), and `launcher.go` re-exports `domains/ec2/systeminstance` launch types as aliases. These clear when ELBv2 is extracted and each edge becomes a consumer-owned capability, which ADR-0001 sequences after the ELBv2 data-plane boundary is settled.
 - `docs/service-interfaces.yaml` (ECS coverage gap, closed): the traced `ecs.bus.*` family and the ECS AWS-API action subjects are both listed. The action subjects are declared as `ecs.*`, the same wildcard convention as `rds.*`: `handlers/ecs/service_nats.go` publishes them from the gateway and `daemon/daemon.go` subscribes to them. Declaring `ecs.bus.*` alone had made `ecs` a managed prefix, so `manifest-lint` failed CI on all 34 undeclared action subjects until this was added.
 - `contracts/ecs/v1` (dead routes): four of five bus subjects have no producer. The scheduler leader still subscribes to `instance-register`, `instance-heartbeat` and `task-state`, which can never fire, and `assign` has neither producer nor consumer; only `service-reconcile` is live. Removal condition: an owner decision to retire these routes (a successor contract under the compatibility rules) or to adopt them in the agent.
@@ -650,13 +651,13 @@ series above; "needs decision" is open.
 
 | Legacy path | What remains | What blocks it |
 |---|---|---|
-| `handlers/iam` | Private identity state and credential issuance; consumed by the EC2 instance and spot-instance AWS adapters, `handlers/sts`, `gateway/sts`, `gateway/bedrock`, the ELBv2 system-instance role path, ECR/Ochre signing-key crypto and `operator/cli`. | Deferred: the ADR's `handlers/iam` row states a shared IAM contract is not inferred from current KV readers; each cross-domain reader needs an IAM capability first. |
-| `handlers/sts` | Everything except the EKS OIDC JWKS lookup, which now reaches `handlers/eks` only through the injected `ClusterJWKSResolver`. | Deferred: per-consumer IAM capabilities (dependency-rule debt above); `handlers/sts` still imports `handlers/iam` directly. |
+| `handlers/iam` | — | Done. The implementation moved to `domains/iam` (package `iam`) in `spinifex#1249` with no API, NATS subject, storage or behaviour change. ADR-0001's row says a shared IAM contract is not inferred from current KV readers, and the move adds none. The cross-domain readers (the EC2 instance and spot-instance AWS adapters, `handlers/sts`, `gateway/sts`, `gateway/bedrock`, the ELBv2 system-instance role path and ECR/Ochre signing-key crypto) still import the implementation and each still needs an IAM capability; that is recorded dependency-rule debt, with its allowlist entries now targeting `domains/iam`, not a prerequisite to relocating IAM's implementation. |
+| `handlers/sts` | Everything except the EKS OIDC JWKS lookup, which now reaches `handlers/eks` only through the injected `ClusterJWKSResolver`. | Deferred: per-consumer IAM capabilities (dependency-rule debt above); `handlers/sts` still imports `domains/iam` directly. |
 | `handlers/ecs` | Task, service and capacity resources (`taskdefinition` already extracted to `domains/ecs/taskdefinition`). | Deferred: ADR-0004 blocked-dependency list — `iam:PassRole` in the gateway adapter and EC2/ELBv2/EIP adapters called as internal APIs. |
 | `handlers/eks` | Nodegroup and cluster resources (`access` and `addon` already extracted). | Deferred: ADR-0004 blocked-dependency list — IAM role/instance-profile creation, Bluebottle `auth`, the NATS CA, and EKS identity (token review, OIDC). |
 | `handlers/rds` | The RDS instance resource (subnet group, parameter group and engine catalogue already extracted). | Deferred: ADR-0004 blocked-dependency list — quota admission, the duplicated expected-node fan-out, the IAM ensurer and the Bluebottle master key. |
 | `handlers/elbv2` | Everything; not yet touched. | Deferred: ADR-0001 sequences this after the ELBv2 data-plane boundary is settled. Needs decision: what evidence establishes that boundary is settled. |
-| `handlers/iam/mock` | A test double for `handlers/iam`. | Moves with `handlers/iam`; it has no independent target. |
+| `handlers/iam/mock` | — | Done. Moved with the implementation to `domains/iam/mock`. |
 | `gateway` (root) | Generic HTTP route assembly, SigV4/registry authentication, the policy gate, request-audit middleware, request-context keys, and the legacy per-service dispatch switch for every service not yet registered. | Deferred: the ADR-0002 phases above (auth/IAM and request-context boundary design) gate moving the audit, context-key and legacy-dispatch mechanics. |
 | `gateway/bedrock` | The Ochre domain core, its stores and the inference engine, together with the Bedrock action layer, in one package with internal cycles. | Deferred: needs the split plan recorded above — domain core and stores to `domains/ochre/<resource>`, inference to an Ochre-owned package, and only the action layer to `domains/ochre/awsapi`. |
 | `gateway/{ecs,eks,rds,iam,sts,elbv2}` | Each service's AWS action adapters and dispatch, not yet registered through ADR-0002. | Deferred: the ADR-0002 registration contract has only reached ECR and ACM so far; each service needs its own registration slice. |
@@ -682,9 +683,9 @@ ADR-0001 completion is structural, so none of these is cleared by remediating se
 | Explicitly deferred by ADR-0001 | `daemon` (Q-28), `lbagent` (Q-30) | Closed for this refactor; the ADR names the prerequisite. |
 | No missing boundary (candidate moves) | `handlers/ecs/bus` → `contracts/ecs/v1` (`a130f6b8e`); `admin/images_promote.go` → `domains/ec2/image` (`263c58167`); `admin` bootstrap account identifiers → `foundation/aws/identifiers` (`3fdaa54d9`); `admin` credential-format generators → `foundation/aws/credentials` `0368e861a`, retiring `admin` | Done. |
 | One narrow missing boundary | Done: `handlers/sts` EKS JWKS lookup → `handlers/eks` `ClusterJWKSResolver` (`254d898a3`); `gateway/tagging` → `domains/tagging` with AWSGW-composed listers (`0d830e901`, `f3545e3ea`, `aedd0b2c4`). Open: `gateway/spx` (owner of operator-plane dispatch registration) | Each needs one decision or one capability, then a mechanical move. |
-| Broader boundary or follows another package | `handlers/iam` and `handlers/iam/mock` (per-consumer IAM capabilities); `handlers/{ecs,eks,rds}` (ADR-0004 blocked-dependency lists); `handlers/elbv2` (data-plane boundary evidence); `accountteardown` (domain cleanup capabilities); `gateway` root (ADR-0002 auth and request-context phase); `gateway/{iam,sts,ecs,eks,rds,elbv2}` (each follows its `handlers/<svc>`, because a new-path `awsapi` cannot import a legacy handler); `gateway/bedrock` (the recorded Ochre split plan) | Recorded debt with removal conditions above. |
+| Broader boundary or follows another package | `handlers/{ecs,eks,rds}` (ADR-0004 blocked-dependency lists); `handlers/elbv2` (data-plane boundary evidence); `accountteardown` (domain cleanup capabilities); `gateway` root (ADR-0002 auth and request-context phase); `gateway/{iam,sts,ecs,eks,rds,elbv2}` (each follows its `handlers/<svc>`, because a new-path `awsapi` cannot import a legacy handler); `gateway/bedrock` (the recorded Ochre split plan) | Recorded debt with removal conditions above. |
 
-Per-consumer IAM capabilities are the widest unblocker: they gate `handlers/iam` itself, STS's IAM reads, Ochre's IAM reads, the ELBv2 role path and the ECS, EKS and RDS ensurers.
+Per-consumer IAM capabilities are the widest unblocker: they gate STS's IAM reads, Ochre's IAM reads, the ELBv2 role path and the ECS, EKS and RDS ensurers.
 
 `runtime/roles/awsgw` carries two recorded residuals from its own move (`931d86d5c`
 above) that are not yet re-homed: `runQuotaReconcile`/`quotaKeyFor` belong to

docs/service-interfaces.yaml

@@ -76,7 +76,7 @@ services:
       - spinifex/domains/ec2/vpc
       - spinifex/handlers/elbv2
       - spinifex/domains/acm
-      - spinifex/handlers/iam
+      - spinifex/domains/iam
       - spinifex/handlers/ecs
       - spinifex/handlers/rds
       - spinifex/domains/rds/subnetgroup

spinifex/accountteardown/cluster.go

@@ -6,8 +6,8 @@ import (
 	"log/slog"
 
 	"github.com/mulgadc/spinifex/spinifex/domains/admission/quota"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvutil"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 	"github.com/nats-io/nats.go/jetstream"
 )
@@ -22,7 +22,7 @@ func NewClusterEngine(
 	nc *nats.Conn,
 	expectedNodes int,
 	nodeIDs []string,
-	svc handlers_iam.IAMService,
+	svc iamdomain.IAMService,
 	buckets BucketStore,
 ) (*Engine, error) {
 	// Refused rather than skipped. A teardown missing its bucket reaper still
@@ -37,7 +37,7 @@ func NewClusterEngine(
 		return nil, fmt.Errorf("get JetStream context: %w", err)
 	}
 
-	names, err := handlers_iam.NewAccountNameIndex(ctx, js)
+	names, err := iamdomain.NewAccountNameIndex(ctx, js)
 	if err != nil {
 		return nil, fmt.Errorf("open account name index: %w", err)
 	}

spinifex/accountteardown/cluster_test.go

@@ -7,7 +7,7 @@ import (
 	"testing"
 
 	"github.com/mulgadc/spinifex/internal/testkit"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
@@ -18,9 +18,9 @@ import (
 func TestNewClusterEngineCoversEveryStage(t *testing.T) {
 	_, nc, _ := testutil.StartTestJetStream(t)
 
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
-	svc, err := handlers_iam.NewIAMServiceImpl(t.Context(), nc, masterKey)
+	svc, err := iamdomain.NewIAMServiceImpl(t.Context(), nc, masterKey)
 	require.NoError(t, err)
 
 	engine, err := NewClusterEngine(t.Context(), nc, 1, []string{"node-1"}, svc, newFakeBuckets())
@@ -47,9 +47,9 @@ func TestNewClusterEngineCoversEveryStage(t *testing.T) {
 func TestNewClusterEngineRegistersTheEC2SettingsSealReaper(t *testing.T) {
 	_, nc, _ := testutil.StartTestJetStream(t)
 
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
-	svc, err := handlers_iam.NewIAMServiceImpl(t.Context(), nc, masterKey)
+	svc, err := iamdomain.NewIAMServiceImpl(t.Context(), nc, masterKey)
 	require.NoError(t, err)
 
 	engine, err := NewClusterEngine(t.Context(), nc, 1, []string{"node-1"}, svc, newFakeBuckets())
@@ -67,9 +67,9 @@ func TestNewClusterEngineRegistersTheEC2SettingsSealReaper(t *testing.T) {
 func TestClusterEnginePrechecks(t *testing.T) {
 	_, nc, _ := testutil.StartTestJetStream(t)
 
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
-	svc, err := handlers_iam.NewIAMServiceImpl(t.Context(), nc, masterKey)
+	svc, err := iamdomain.NewIAMServiceImpl(t.Context(), nc, masterKey)
 	require.NoError(t, err)
 
 	engine, err := NewClusterEngine(t.Context(), nc, 1, []string{"node-1"}, svc, newFakeBuckets())

spinifex/accountteardown/ec2_account_settings_seal_capability.go

@@ -3,7 +3,6 @@ package accountteardown
 import (
 	"context"
 	"encoding/json"
-	"errors"
 	"sort"
 	"strings"
 	"sync"
@@ -184,8 +183,8 @@ func errSealCapabilityIncomplete(report sealCapabilityReport) error {
 	if len(parts) > 0 {
 		detail = strings.Join(parts, ", ")
 	}
-	return errors.Join(
+	return awserrors.WithRetryAfter(
 		awserrors.Errorf(awserrors.ErrorServiceUnavailable, "EC2 account-settings seal capability incomplete: %s", detail),
-		awserrors.RetryAfter(awserrors.ErrorServiceUnavailable, 5*time.Second),
+		5*time.Second,
 	)
 }

spinifex/accountteardown/ec2_account_settings_seal_capability_test.go

@@ -297,3 +297,15 @@ func TestSealCapabilityBoundedByOneDeadline(t *testing.T) {
 	assert.False(t, report.Complete)
 	assert.Less(t, elapsed, client.wait+time.Second, "five silent nodes must not cost five deadlines")
 }
+
+// The stuck reason is logged and stored as one line naming each failing node;
+// the retry hint and the ServiceUnavailable code travel with it.
+func TestSealCapabilityIncompleteErrorIsOneLine(t *testing.T) {
+	err := errSealCapabilityIncomplete(sealCapabilityReport{Nodes: []nodeSealCapability{{Node: "node-1", Reason: sealCapabilityReasonDuplicate}}})
+
+	assert.Equal(t, "EC2 account-settings seal capability incomplete: node-1=duplicate: ServiceUnavailable", err.Error())
+	d, ok := awserrors.ResolveRetryAfter(err)
+	assert.True(t, ok)
+	assert.Equal(t, 5*time.Second, d)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorServiceUnavailable))
+}

spinifex/accountteardown/ec2_platform_inventory_test.go

@@ -17,9 +17,9 @@ import (
 	ec2v1 "github.com/mulgadc/spinifex/contracts/ec2/v1"
 	"github.com/mulgadc/spinifex/internal/testkit"
 	"github.com/mulgadc/spinifex/spinifex/domains/ec2/accountinventory"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	awserrors "github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
 	natsmsg "github.com/mulgadc/spinifex/spinifex/foundation/messaging/nats"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -165,9 +165,9 @@ func TestEC2PlatformInventoryClientDistrustsAMalformedReply(t *testing.T) {
 // a Teardown request marked dry run and a full teardown all leave it uncalled.
 func TestOnlyTheExplicitDryRunCallsTheEC2PlatformInventory(t *testing.T) {
 	_, nc, _ := testutil.StartTestJetStream(t)
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
-	svc, err := handlers_iam.NewIAMServiceImpl(t.Context(), nc, masterKey)
+	svc, err := iamdomain.NewIAMServiceImpl(t.Context(), nc, masterKey)
 	require.NoError(t, err)
 	cluster, err := NewClusterEngine(t.Context(), nc, 1, []string{"node-1"}, svc, newFakeBuckets())
 	require.NoError(t, err)

spinifex/accountteardown/engine.go

@@ -10,7 +10,7 @@ import (
 	"time"
 )
 
-// AccountStatusTerminating mirrors handlers_iam.AccountStatusTerminating.
+// AccountStatusTerminating mirrors iamdomain.AccountStatusTerminating.
 // Redeclared so this package does not depend on the IAM service to be tested.
 const AccountStatusTerminating = "TERMINATING"
 

spinifex/accountteardown/engine_test.go

@@ -13,7 +13,7 @@ import (
 	"testing"
 	"time"
 
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 )
 
 const (
@@ -129,11 +129,11 @@ func TestProtectedAccountsMatchTheRestOfTheSystem(t *testing.T) {
 		if _, protected := protectedAccountIDs[id]; !protected {
 			t.Fatalf("account %s is not protected by teardown", id)
 		}
-		if _, protected := handlers_iam.UndeletableAccountIDs[id]; !protected {
+		if _, protected := iamdomain.UndeletableAccountIDs[id]; !protected {
 			t.Fatalf("account %s is not protected by the IAM service", id)
 		}
 	}
-	if len(protectedAccountIDs) != len(handlers_iam.UndeletableAccountIDs) {
+	if len(protectedAccountIDs) != len(iamdomain.UndeletableAccountIDs) {
 		t.Fatal("teardown and the IAM service disagree on which accounts are protected")
 	}
 }

spinifex/accountteardown/reapers_account.go

@@ -5,13 +5,13 @@ import (
 	"errors"
 	"fmt"
 
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/nats-io/nats.go/jetstream"
 )
 
 // AccountReapers returns the final-stage reapers: the per-account counters and
 // reservations that outlive every resource but must not outlive the account.
-func AccountReapers(svc handlers_iam.IAMService, names *handlers_iam.AccountNameIndex, usage jetstream.KeyValue) []Reaper {
+func AccountReapers(svc iamdomain.IAMService, names *iamdomain.AccountNameIndex, usage jetstream.KeyValue) []Reaper {
 	return []Reaper{
 		&quotaUsageReaper{usage: usage},
 		&nameReservationReaper{svc: svc, names: names},
@@ -53,8 +53,8 @@ func (r *quotaUsageReaper) Delete(ctx context.Context, _ string, resource Resour
 // up again. A customer who deletes an account and returns must not be told the
 // address is taken by a tenant that no longer exists.
 type nameReservationReaper struct {
-	svc   handlers_iam.IAMService
-	names *handlers_iam.AccountNameIndex
+	svc   iamdomain.IAMService
+	names *iamdomain.AccountNameIndex
 }
 
 func (r *nameReservationReaper) Kind() string { return "name-reservation" }
@@ -92,12 +92,12 @@ func (r *nameReservationReaper) Delete(ctx context.Context, accountID string, re
 // IAMAccounts adapts the IAM service to the engine's AccountStore. The engine
 // declares its own Account type so its tests need no cluster; this is the one
 // place the two meet.
-type IAMAccounts struct{ svc handlers_iam.IAMService }
+type IAMAccounts struct{ svc iamdomain.IAMService }
 
 var _ AccountStore = (*IAMAccounts)(nil)
 
 // NewIAMAccounts wraps svc as the engine's AccountStore.
-func NewIAMAccounts(svc handlers_iam.IAMService) *IAMAccounts {
+func NewIAMAccounts(svc iamdomain.IAMService) *IAMAccounts {
 	return &IAMAccounts{svc: svc}
 }
 

spinifex/accountteardown/reapers_account_test.go

@@ -10,8 +10,8 @@ import (
 	"testing"
 
 	"github.com/mulgadc/spinifex/internal/testkit"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvutil"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go/jetstream"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -22,20 +22,20 @@ import (
 type accountIAM struct {
 	fakeIAM
 
-	account   *handlers_iam.Account
+	account   *iamdomain.Account
 	getErr    error
 	deleted   bool
 	setStatus string
 }
 
-func (f *accountIAM) GetAccount(string) (*handlers_iam.Account, error) {
+func (f *accountIAM) GetAccount(string) (*iamdomain.Account, error) {
 	if f.getErr != nil {
 		return nil, f.getErr
 	}
 	return f.account, nil
 }
 
-func (f *accountIAM) SetAccountStatus(_, status string) (*handlers_iam.Account, error) {
+func (f *accountIAM) SetAccountStatus(_, status string) (*iamdomain.Account, error) {
 	f.setStatus = status
 	updated := *f.account
 	updated.Status = status
@@ -102,12 +102,12 @@ func TestNameReservationReaperReleasesTheName(t *testing.T) {
 	_, js, _ := testKV(t, "spinifex-name-reaper-warmup")
 	ctx := testCtx(t)
 
-	names, err := handlers_iam.NewAccountNameIndex(ctx, js)
+	names, err := iamdomain.NewAccountNameIndex(ctx, js)
 	require.NoError(t, err)
 	require.NoError(t, names.Reserve(ctx, "tenant@example.com", "token-1"))
 	require.NoError(t, names.Commit(ctx, "tenant@example.com", "000000000042", "token-1"))
 
-	svc := &accountIAM{account: &handlers_iam.Account{
+	svc := &accountIAM{account: &iamdomain.Account{
 		AccountID: "000000000042", AccountName: "tenant@example.com",
 	}}
 	reaper := &nameReservationReaper{svc: svc, names: names}
@@ -130,12 +130,12 @@ func TestNameReservationReaperIgnoresAnotherAccountsName(t *testing.T) {
 	_, js, _ := testKV(t, "spinifex-name-reaper-other")
 	ctx := testCtx(t)
 
-	names, err := handlers_iam.NewAccountNameIndex(ctx, js)
+	names, err := iamdomain.NewAccountNameIndex(ctx, js)
 	require.NoError(t, err)
 	require.NoError(t, names.Reserve(ctx, "tenant@example.com", "token-1"))
 	require.NoError(t, names.Commit(ctx, "tenant@example.com", "000000000099", "token-1"))
 
-	svc := &accountIAM{account: &handlers_iam.Account{
+	svc := &accountIAM{account: &iamdomain.Account{
 		AccountID: "000000000042", AccountName: "tenant@example.com",
 	}}
 
@@ -154,7 +154,7 @@ func TestNameReservationReaperReportsAMissingAccount(t *testing.T) {
 	_, js, _ := testKV(t, "spinifex-name-reaper-missing")
 	ctx := testCtx(t)
 
-	names, err := handlers_iam.NewAccountNameIndex(ctx, js)
+	names, err := iamdomain.NewAccountNameIndex(ctx, js)
 	require.NoError(t, err)
 
 	svc := &accountIAM{getErr: errors.New("account not found")}
@@ -177,10 +177,10 @@ func TestAccountReapersRunLast(t *testing.T) {
 // The adapter is the one place the engine's Account and the IAM record meet,
 // so a field dropped in translation would silently break the name confirmation.
 func TestIAMAccountsAdapterCarriesEveryField(t *testing.T) {
-	svc := &accountIAM{account: &handlers_iam.Account{
+	svc := &accountIAM{account: &iamdomain.Account{
 		AccountID:   "000000000042",
 		AccountName: "tenant@example.com",
-		Status:      handlers_iam.AccountStatusActive,
+		Status:      iamdomain.AccountStatusActive,
 	}}
 	accounts := NewIAMAccounts(svc)
 
@@ -188,12 +188,12 @@ func TestIAMAccountsAdapterCarriesEveryField(t *testing.T) {
 	require.NoError(t, err)
 	assert.Equal(t, "000000000042", account.AccountID)
 	assert.Equal(t, "tenant@example.com", account.AccountName)
-	assert.Equal(t, handlers_iam.AccountStatusActive, account.Status)
+	assert.Equal(t, iamdomain.AccountStatusActive, account.Status)
 
 	updated, err := accounts.SetAccountStatus("000000000042", AccountStatusTerminating)
 	require.NoError(t, err)
 	assert.Equal(t, AccountStatusTerminating, updated.Status)
-	assert.Equal(t, handlers_iam.AccountStatusTerminating, svc.setStatus)
+	assert.Equal(t, iamdomain.AccountStatusTerminating, svc.setStatus)
 
 	require.NoError(t, accounts.DeleteAccount("000000000042"))
 	assert.True(t, svc.deleted)
@@ -203,5 +203,5 @@ func TestIAMAccountsAdapterCarriesEveryField(t *testing.T) {
 // two must not drift: a mismatch would mark accounts with a status the auth
 // gate does not recognise, and they would keep authenticating.
 func TestTerminatingStatusMatchesTheIAMRecord(t *testing.T) {
-	assert.Equal(t, handlers_iam.AccountStatusTerminating, AccountStatusTerminating)
+	assert.Equal(t, iamdomain.AccountStatusTerminating, AccountStatusTerminating)
 }

spinifex/accountteardown/reapers_iam.go

@@ -6,7 +6,7 @@ import (
 
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/iam"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 )
 
 // IAMReapers returns the identity-stage reapers in teardown order.
@@ -14,7 +14,7 @@ import (
 // Access keys go first and deliberately: it is the second quiesce, and unlike
 // the account-status gate it holds even against a caller that reached the
 // cluster some other way.
-func IAMReapers(svc handlers_iam.IAMService) []Reaper {
+func IAMReapers(svc iamdomain.IAMService) []Reaper {
 	return []Reaper{
 		&accessKeyReaper{svc: svc},
 		&instanceProfileReaper{svc: svc},
@@ -34,7 +34,7 @@ func IAMReapers(svc handlers_iam.IAMService) []Reaper {
 // They survive their cluster's deletion, so nothing else in teardown reaches
 // them and an account would be left holding a trust anchor for a cluster that
 // no longer exists.
-type oidcProviderReaper struct{ svc handlers_iam.IAMService }
+type oidcProviderReaper struct{ svc iamdomain.IAMService }
 
 func (r *oidcProviderReaper) Kind() string { return "oidc-provider" }
 func (r *oidcProviderReaper) Stage() Stage { return StagePlatform }
@@ -62,7 +62,7 @@ func (r *oidcProviderReaper) Delete(_ context.Context, accountID string, resourc
 	return ignoreAlreadyGone(err)
 }
 
-type accessKeyReaper struct{ svc handlers_iam.IAMService }
+type accessKeyReaper struct{ svc iamdomain.IAMService }
 
 func (r *accessKeyReaper) Kind() string { return "access-key" }
 func (r *accessKeyReaper) Stage() Stage { return StageIdentity }
@@ -102,7 +102,7 @@ func (r *accessKeyReaper) Delete(_ context.Context, accountID string, resource R
 	return ignoreAlreadyGone(err)
 }
 
-type iamUserReaper struct{ svc handlers_iam.IAMService }
+type iamUserReaper struct{ svc iamdomain.IAMService }
 
 func (r *iamUserReaper) Kind() string { return "iam-user" }
 func (r *iamUserReaper) Stage() Stage { return StageIdentity }
@@ -166,7 +166,7 @@ func (r *iamUserReaper) Delete(_ context.Context, accountID string, resource Res
 	return ignoreAlreadyGone(err)
 }
 
-type iamRoleReaper struct{ svc handlers_iam.IAMService }
+type iamRoleReaper struct{ svc iamdomain.IAMService }
 
 func (r *iamRoleReaper) Kind() string { return "iam-role" }
 func (r *iamRoleReaper) Stage() Stage { return StageIdentity }
@@ -228,7 +228,7 @@ func (r *iamRoleReaper) Delete(_ context.Context, accountID string, resource Res
 	return ignoreAlreadyGone(err)
 }
 
-type iamGroupReaper struct{ svc handlers_iam.IAMService }
+type iamGroupReaper struct{ svc iamdomain.IAMService }
 
 func (r *iamGroupReaper) Kind() string { return "iam-group" }
 func (r *iamGroupReaper) Stage() Stage { return StageIdentity }
@@ -254,7 +254,7 @@ func (r *iamGroupReaper) Delete(_ context.Context, accountID string, resource Re
 	return ignoreAlreadyGone(err)
 }
 
-type iamPolicyReaper struct{ svc handlers_iam.IAMService }
+type iamPolicyReaper struct{ svc iamdomain.IAMService }
 
 func (r *iamPolicyReaper) Kind() string { return "iam-policy" }
 func (r *iamPolicyReaper) Stage() Stage { return StageIdentity }
@@ -302,7 +302,7 @@ func (r *iamPolicyReaper) Delete(_ context.Context, accountID string, resource R
 	return ignoreAlreadyGone(err)
 }
 
-type instanceProfileReaper struct{ svc handlers_iam.IAMService }
+type instanceProfileReaper struct{ svc iamdomain.IAMService }
 
 func (r *instanceProfileReaper) Kind() string { return "instance-profile" }
 func (r *instanceProfileReaper) Stage() Stage { return StageIdentity }

spinifex/accountteardown/reapers_iam_test.go

@@ -11,7 +11,7 @@ import (
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/iam"
 	"github.com/mulgadc/spinifex/internal/testkit"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
@@ -21,7 +21,7 @@ import (
 // stay unimplemented — calling one is a test failure by panic, which is the
 // honest outcome for a reaper reaching somewhere it was not meant to.
 type fakeIAM struct {
-	handlers_iam.IAMService
+	iamdomain.IAMService
 
 	calls []string
 
@@ -396,9 +396,9 @@ func TestPolicyReaperAsksOnlyForLocalPolicies(t *testing.T) {
 // DeletePolicy refuses it until they are gone.
 func TestPolicyReaperDeletesAPolicyWithSeveralVersions(t *testing.T) {
 	_, nc, _ := testutil.StartTestJetStream(t)
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
-	svc, err := handlers_iam.NewIAMServiceImpl(t.Context(), nc, masterKey)
+	svc, err := iamdomain.NewIAMServiceImpl(t.Context(), nc, masterKey)
 	require.NoError(t, err)
 
 	const accountID = "000000000042"

spinifex/accountteardown/reapers_quota_override_current_behaviour_test.go

@@ -7,8 +7,8 @@ import (
 
 	"github.com/mulgadc/spinifex/internal/testkit"
 	"github.com/mulgadc/spinifex/spinifex/domains/admission/quota"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvutil"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go/jetstream"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -36,7 +36,7 @@ func TestCurrentBehaviour_AccountTermination_QuotaOverridesSurviveTeardown(t *te
 	require.NoError(t, svc.PutAccountQuota(ctx, testAccountID, quota.Overrides{VCPUs: &unlimited}, "operator"))
 
 	accounts := newFakeAccounts()
-	iam := &accountIAM{account: &handlers_iam.Account{AccountID: testAccountID, AccountName: testAccountName}}
+	iam := &accountIAM{account: &iamdomain.Account{AccountID: testAccountID, AccountName: testAccountName}}
 	engine := quietEngine(accounts, AccountReapers(iam, nil, usage)...)
 
 	result, err := engine.Teardown(ctx, request())

spinifex/daemon/daemon.go

@@ -63,6 +63,7 @@ import (
 	ec2volume "github.com/mulgadc/spinifex/spinifex/domains/ec2/volume"
 	ec2vpc "github.com/mulgadc/spinifex/spinifex/domains/ec2/vpc"
 	"github.com/mulgadc/spinifex/spinifex/domains/ecr"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/domains/network/external"
 	"github.com/mulgadc/spinifex/spinifex/domains/network/external/dhcp"
 	"github.com/mulgadc/spinifex/spinifex/domains/network/external/exonet"
@@ -78,7 +79,6 @@ import (
 	handlers_ecs "github.com/mulgadc/spinifex/spinifex/handlers/ecs"
 	handlers_eks "github.com/mulgadc/spinifex/spinifex/handlers/eks"
 	handlers_elbv2 "github.com/mulgadc/spinifex/spinifex/handlers/elbv2"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	handlers_rds "github.com/mulgadc/spinifex/spinifex/handlers/rds"
 	"github.com/mulgadc/spinifex/spinifex/providers/ebs"
 	"github.com/mulgadc/spinifex/spinifex/providers/objectstore"
@@ -297,7 +297,7 @@ type Daemon struct {
 
 	// iamEnsurerMu guards the lazily-built system-role IAM service (systemRoleEnsurer).
 	iamEnsurerMu     sync.Mutex
-	iamEnsurerCached handlers_iam.SystemInstanceRoleEnsurer
+	iamEnsurerCached iamdomain.SystemInstanceRoleEnsurer
 
 	mu sync.Mutex
 }
@@ -2006,8 +2006,8 @@ func (d *Daemon) startCluster() error {
 	if err != nil {
 		return fmt.Errorf("jetstream handle for account status reader: %w", err)
 	}
-	var _ ec2account.AccountTerminationChecker = (*handlers_iam.AccountStatusReader)(nil)
-	d.accountService.SetAccountTerminationChecker(handlers_iam.NewAccountStatusReader(accountStatusJS))
+	var _ ec2account.AccountTerminationChecker = (*iamdomain.AccountStatusReader)(nil)
+	d.accountService.SetAccountTerminationChecker(iamdomain.NewAccountStatusReader(accountStatusJS))
 
 	// ELBv2 resolves listener certificate private keys through its own ACM
 	// Store over the same JetStream bucket the ACM service writes, so it needs

spinifex/daemon/daemon_system_instance_test.go

@@ -16,11 +16,11 @@ import (
 	ec2eip "github.com/mulgadc/spinifex/spinifex/domains/ec2/eip"
 	"github.com/mulgadc/spinifex/spinifex/domains/ec2/systeminstance"
 	ec2vpc "github.com/mulgadc/spinifex/spinifex/domains/ec2/vpc"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/domains/network/external"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/tags"
 	"github.com/mulgadc/spinifex/spinifex/handlers/elbv2"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/mulgadc/spinifex/spinifex/runtime/compute/vm"
 	"github.com/nats-io/nats.go"
 	"github.com/stretchr/testify/assert"
@@ -623,7 +623,7 @@ func newRefreshTestDaemon(t *testing.T) (*Daemon, *nats.Conn, string) {
 	tmpDir := t.TempDir()
 	t.Setenv("XDG_RUNTIME_DIR", tmpDir)
 	_, nc, _ := testutil.StartTestJetStream(t)
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
 	svc, err := handlers_elbv2.NewELBv2ServiceImplWithNATS(nil, nc, masterKey)
 	require.NoError(t, err)

spinifex/daemon/daemon_wire_lb_test.go

@@ -7,8 +7,8 @@ import (
 	"github.com/mulgadc/spinifex/internal/testkit"
 	"github.com/mulgadc/spinifex/spinifex/bootstrap/config"
 	ec2eip "github.com/mulgadc/spinifex/spinifex/domains/ec2/eip"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	handlers_elbv2 "github.com/mulgadc/spinifex/spinifex/handlers/elbv2"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -20,7 +20,7 @@ func newWireLBTestDaemon(t *testing.T, cfg *config.Config) *Daemon {
 	t.Helper()
 	_, nc, _ := testutil.StartTestJetStream(t)
 
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
 	svc, err := handlers_elbv2.NewELBv2ServiceImplWithNATS(nil, nc, masterKey)
 	require.NoError(t, err)
@@ -203,7 +203,7 @@ func TestWireLBAgentConfig_GatewayURL_SingleNodeBindMatchesAdvertise(t *testing.
 // newSubscribeTestDaemon creates a minimal Daemon wired enough for subscribeAll.
 func newSubscribeTestDaemon(t *testing.T, nc *nats.Conn, gatewayURL, accessKey string) *Daemon {
 	t.Helper()
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
 	svc, err := handlers_elbv2.NewELBv2ServiceImplWithNATS(nil, nc, masterKey)
 	require.NoError(t, err)

spinifex/daemon/dns_reconcile_test.go

@@ -9,10 +9,10 @@ import (
 	"testing"
 
 	"github.com/mulgadc/spinifex/internal/testkit"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvstore"
 	handlers_eks "github.com/mulgadc/spinifex/spinifex/handlers/eks"
 	handlers_elbv2 "github.com/mulgadc/spinifex/spinifex/handlers/elbv2"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	handlers_rds "github.com/mulgadc/spinifex/spinifex/handlers/rds"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -81,7 +81,7 @@ func TestElbv2WatchBuckets_NoStoreIsNilSafe(t *testing.T) {
 
 func TestElbv2WatchBuckets_WithStoreReturnsWatchBucket(t *testing.T) {
 	_, nc, _ := testutil.StartTestJetStream(t)
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
 	svc, err := handlers_elbv2.NewELBv2ServiceImplWithNATS(nil, nc, masterKey)
 	require.NoError(t, err)

spinifex/daemon/ecs_deps.go

@@ -6,8 +6,8 @@ import (
 	"path/filepath"
 
 	"github.com/mulgadc/bluebottle/pkg/masterkey"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	handlers_ecs "github.com/mulgadc/spinifex/spinifex/handlers/ecs"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 )
 
 // buildECSServiceDeps assembles the ECS service Deps for ProvisionCapacity: the
@@ -47,8 +47,8 @@ func (d *Daemon) buildECSServiceDeps() handlers_ecs.Deps {
 	// to have responders. A daemon that started first got one attempt, failed
 	// it, and had capacity provisioning off for the life of the process with no
 	// symptom but a 500 from ProvisionCapacity.
-	iamSvc, iamErr := initServiceWithRetry("ECS IAM service", func() (*handlers_iam.IAMServiceImpl, error) {
-		return handlers_iam.NewIAMServiceImpl(d.ctx, d.natsConn, masterKey)
+	iamSvc, iamErr := initServiceWithRetry("ECS IAM service", func() (*iamdomain.IAMServiceImpl, error) {
+		return iamdomain.NewIAMServiceImpl(d.ctx, d.natsConn, masterKey)
 	})
 	if iamErr != nil {
 		// Off rather than fatal: ECS CRUD, EKS and EC2 do not need this, and a

spinifex/daemon/eks_deps.go

@@ -10,8 +10,8 @@ import (
 	"github.com/mulgadc/bluebottle/pkg/masterkey"
 	"github.com/mulgadc/spinifex/spinifex/bootstrap/config"
 	ec2placementgroup "github.com/mulgadc/spinifex/spinifex/domains/ec2/placementgroup"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	handlers_eks "github.com/mulgadc/spinifex/spinifex/handlers/eks"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/mulgadc/spinifex/spinifex/providers/objectstore"
 )
 
@@ -24,7 +24,7 @@ import (
 // first-use (an LB/cluster create) the backend is up. Returns nil (caller falls
 // back to static creds) until a build succeeds; retried on each call until then,
 // cached once it works.
-func (d *Daemon) systemRoleEnsurer() handlers_iam.SystemInstanceRoleEnsurer {
+func (d *Daemon) systemRoleEnsurer() iamdomain.SystemInstanceRoleEnsurer {
 	d.iamEnsurerMu.Lock()
 	defer d.iamEnsurerMu.Unlock()
 	if d.iamEnsurerCached != nil {
@@ -36,7 +36,7 @@ func (d *Daemon) systemRoleEnsurer() handlers_iam.SystemInstanceRoleEnsurer {
 			"err", err)
 		return nil
 	}
-	iamSvc, iamErr := handlers_iam.NewIAMServiceImpl(d.ctx, d.natsConn, masterKey)
+	iamSvc, iamErr := iamdomain.NewIAMServiceImpl(d.ctx, d.natsConn, masterKey)
 	if iamErr != nil {
 		slog.Warn("System role ensurer: IAM service init failed (retried on next launch); system VMs fall back to baked static creds",
 			"err", iamErr)

spinifex/domains/acm/service_impl_test.go

@@ -21,9 +21,9 @@ import (
 	"github.com/aws/aws-sdk-go/private/protocol/json/jsonutil"
 	"github.com/aws/aws-sdk-go/service/acm"
 	"github.com/mulgadc/spinifex/internal/testkit"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/arn"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
@@ -33,7 +33,7 @@ const testAccountID = "000000000001"
 func setupACMService(t *testing.T) *ACMServiceImpl {
 	t.Helper()
 	_, nc, _ := testutil.StartTestJetStream(t)
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
 	svc, err := NewACMServiceImplWithNATS(t.Context(), nil, nc, masterKey)
 	require.NoError(t, err)
@@ -307,7 +307,7 @@ func TestImportCertificate_PrivateKeyEncryptedAtRest(t *testing.T) {
 // TestDescribeCertificate_DoesNotLeakPrivateKey asserts the public
 // CertificateDetail response can never carry key material, however the
 // record is stored. Precedent: TestSensitiveDataNotLogged_MasterKey in
-// handlers/iam/service_impl_test.go.
+// domains/iam/service_impl_test.go.
 func TestDescribeCertificate_DoesNotLeakPrivateKey(t *testing.T) {
 	t.Parallel()
 	svc := setupACMService(t)
@@ -331,7 +331,7 @@ func TestDescribeCertificate_DoesNotLeakPrivateKey(t *testing.T) {
 // TestSensitiveDataNotLogged_ACMPrivateKey asserts ImportCertificate and
 // DescribeCertificate never place the raw private key PEM in log output.
 // Precedent: TestSensitiveDataNotLogged_MasterKey in
-// handlers/iam/service_impl_test.go.
+// domains/iam/service_impl_test.go.
 // fakeCertAuthority is a minimal CertAuthority for testing RequestCertificate's
 // PRIVATE_CA path without the real tenant CA implementation (domains/acm/privateca.go
 // lives on a separate branch).

spinifex/domains/acm/store.go

@@ -11,10 +11,10 @@ import (
 	"strings"
 	"time"
 
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/arn"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvstore"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvutil"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 	"github.com/nats-io/nats.go/jetstream"
 )
@@ -161,7 +161,7 @@ type CertRecord struct {
 type Store struct {
 	store *kvstore.Store[CertRecord]
 	// masterKey encrypts/decrypts CertRecord.PrivateKey at rest (AES-256-GCM via
-	// handlers_iam.EncryptSecret/DecryptSecret). Every Store over the ACM bucket
+	// iamdomain.EncryptSecret/DecryptSecret). Every Store over the ACM bucket
 	// — the ACM service and ELBv2's independent read-only Store alike — must be
 	// constructed with the same deployment key: a keyed writer and an unkeyed (or
 	// differently keyed) reader of the same bucket disagree silently, with the
@@ -221,7 +221,7 @@ func certKey(certArn string) string {
 // via the legacy-passthrough path in decryptPrivateKey is therefore
 // re-encrypted the next time it is put.
 func (s *Store) PutCert(ctx context.Context, rec *CertRecord) error {
-	ciphertext, err := handlers_iam.EncryptSecret(rec.PrivateKey, s.masterKey)
+	ciphertext, err := iamdomain.EncryptSecret(rec.PrivateKey, s.masterKey)
 	if err != nil {
 		return fmt.Errorf("encrypt private key: %w", err)
 	}
@@ -324,7 +324,7 @@ func isPlaintextPrivateKeyPEM(raw string) bool {
 // ciphertext, garbage) is a hard error rather than a silent plaintext
 // fallback.
 func (s *Store) decryptPrivateKey(rec *CertRecord) error {
-	if plaintext, err := handlers_iam.DecryptSecret(rec.PrivateKey, s.masterKey); err == nil {
+	if plaintext, err := iamdomain.DecryptSecret(rec.PrivateKey, s.masterKey); err == nil {
 		rec.PrivateKey = plaintext
 		return nil
 	}

spinifex/domains/acm/store_test.go

@@ -8,8 +8,8 @@ import (
 	"time"
 
 	"github.com/mulgadc/spinifex/internal/testkit"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvstore"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go/jetstream"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -21,7 +21,7 @@ import (
 func setupACMStore(t *testing.T) *Store {
 	t.Helper()
 	_, nc, _ := testutil.StartTestJetStream(t)
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
 	store, err := NewStore(t.Context(), nc, masterKey)
 	require.NoError(t, err)

spinifex/domains/admission/quota/cleanup.go

@@ -1,93 +0,0 @@
-package quota
-
-import (
-	"context"
-	"errors"
-	"fmt"
-
-	awserrors "github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	awsidentifiers "github.com/mulgadc/spinifex/spinifex/foundation/aws/identifiers"
-	"github.com/nats-io/nats.go/jetstream"
-)
-
-// OverrideCleanupDisposition is CleanupAccount's outcome for an account's
-// quota override key.
-type OverrideCleanupDisposition string
-
-const (
-	OverrideCleanupDeleted OverrideCleanupDisposition = "deleted"
-	OverrideCleanupAbsent  OverrideCleanupDisposition = "absent"
-	OverrideCleanupFailed  OverrideCleanupDisposition = "failed"
-)
-
-// evidenceRevisionGuardedDelete and evidenceObservedAbsent are the fixed
-// evidence bases CleanupAccount reports; see kvstore/bucket.go:112-120 for why
-// a plain read cannot support a stronger absence claim.
-const (
-	evidenceRevisionGuardedDelete = "revision-guarded-delete"
-	evidenceObservedAbsent        = "observed-absent-at-scan"
-)
-
-// cleanupMaxAttempts bounds retries against a revision that keeps moving
-// under concurrent writes, so cleanup fails closed instead of spinning.
-const cleanupMaxAttempts = 5
-
-// OverrideCleanup is CleanupAccount's result: what happened to an account's
-// quota override key, and the fixed evidence basis for that claim.
-type OverrideCleanup struct {
-	Disposition OverrideCleanupDisposition
-	Evidence    string
-	Revision    uint64
-}
-
-// CleanupAccount removes accountID's quota override key so a reused account ID
-// does not inherit a deleted account's raised or lowered limits. It is
-// unwired: nothing calls it yet.
-func (s *Service) CleanupAccount(ctx context.Context, accountID string) (*OverrideCleanup, error) {
-	if accountID == "" {
-		return nil, awserrors.Errorf(awserrors.ErrorMissingParameter, "accountID is required for quota override cleanup")
-	}
-	if !awsidentifiers.IsAccountID(accountID) {
-		return nil, awserrors.Errorf(awserrors.ErrorInvalidParameterValue, "accountID %q is not a valid AWS account ID", accountID)
-	}
-	if accountID == awsidentifiers.GlobalAccountID {
-		return nil, awserrors.Errorf(awserrors.ErrorInvalidParameterValue, "quota override cleanup must not target the global account")
-	}
-	if s == nil || s.overrides == nil {
-		return &OverrideCleanup{Disposition: OverrideCleanupFailed}, errors.New("quota: override bucket not configured")
-	}
-
-	// accountID is never resolved through any fallback here: cleanup must act
-	// on exactly accountID and never fall back to the global key.
-	for range cleanupMaxAttempts {
-		entry, err := s.overrides.Get(ctx, accountID)
-		if err != nil {
-			if errors.Is(err, jetstream.ErrKeyNotFound) {
-				return &OverrideCleanup{Disposition: OverrideCleanupAbsent, Evidence: evidenceObservedAbsent}, nil
-			}
-			return &OverrideCleanup{Disposition: OverrideCleanupFailed}, fmt.Errorf("get quota override for cleanup: %w", err)
-		}
-		rev := entry.Revision()
-
-		err = s.overrides.Delete(ctx, accountID, jetstream.LastRevision(rev))
-		switch {
-		case err == nil:
-			return &OverrideCleanup{
-				Disposition: OverrideCleanupDeleted,
-				Evidence:    evidenceRevisionGuardedDelete,
-				Revision:    rev,
-			}, nil
-		case errors.Is(err, jetstream.ErrKeyNotFound):
-			return &OverrideCleanup{Disposition: OverrideCleanupAbsent, Evidence: evidenceObservedAbsent}, nil
-		case errors.Is(err, jetstream.ErrKeyRevisionMismatch):
-			// A concurrent write landed between the read and the guarded
-			// delete; retry against whatever revision now holds.
-			continue
-		default:
-			return &OverrideCleanup{Disposition: OverrideCleanupFailed}, fmt.Errorf("delete quota override for cleanup: %w", err)
-		}
-	}
-
-	return &OverrideCleanup{Disposition: OverrideCleanupFailed},
-		fmt.Errorf("quota override cleanup for account: exhausted %d attempts against concurrent writes", cleanupMaxAttempts)
-}

spinifex/domains/admission/quota/cleanup_test.go

@@ -1,184 +0,0 @@
-//test:in-package — CleanupAccount deletes through the unexported overrides field, and the retry/race behaviour is only observable from inside the package.
-package quota
-
-import (
-	"context"
-	"errors"
-	"testing"
-
-	awsidentifiers "github.com/mulgadc/spinifex/spinifex/foundation/aws/identifiers"
-	"github.com/nats-io/nats.go/jetstream"
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-)
-
-// overrideCleanupFakeKV wraps a real KeyValue so a test can inject a race
-// between CleanupAccount's read and its guarded delete, or a store failure,
-// deterministically rather than through goroutine timing.
-type overrideCleanupFakeKV struct {
-	jetstream.KeyValue
-
-	deleteCalls  int
-	beforeDelete func(attempt int) error
-	getErr       error
-}
-
-func (f *overrideCleanupFakeKV) Get(ctx context.Context, key string) (jetstream.KeyValueEntry, error) {
-	if f.getErr != nil {
-		return nil, f.getErr
-	}
-	return f.KeyValue.Get(ctx, key)
-}
-
-func (f *overrideCleanupFakeKV) Delete(ctx context.Context, key string, opts ...jetstream.KVDeleteOpt) error {
-	f.deleteCalls++
-	if f.beforeDelete != nil {
-		if err := f.beforeDelete(f.deleteCalls); err != nil {
-			return err
-		}
-	}
-	return f.KeyValue.Delete(ctx, key, opts...)
-}
-
-func TestCleanupAccount_DeletesTheOverrideKey(t *testing.T) {
-	svc, _ := newOverrideService(t)
-	ctx := t.Context()
-	require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(32)}, "operator"))
-
-	result, err := svc.CleanupAccount(ctx, testAccount)
-	require.NoError(t, err)
-	assert.Equal(t, OverrideCleanupDeleted, result.Disposition)
-	assert.Equal(t, evidenceRevisionGuardedDelete, result.Evidence)
-	assert.NotZero(t, result.Revision)
-
-	_, err = svc.overrides.Get(ctx, testAccount)
-	assert.ErrorIs(t, err, jetstream.ErrKeyNotFound)
-}
-
-func TestCleanupAccount_RepeatCallIsIdempotent(t *testing.T) {
-	svc, _ := newOverrideService(t)
-	ctx := t.Context()
-	require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(32)}, "operator"))
-
-	first, err := svc.CleanupAccount(ctx, testAccount)
-	require.NoError(t, err)
-	assert.Equal(t, OverrideCleanupDeleted, first.Disposition)
-
-	second, err := svc.CleanupAccount(ctx, testAccount)
-	require.NoError(t, err)
-	assert.Equal(t, OverrideCleanupAbsent, second.Disposition)
-	assert.Equal(t, evidenceObservedAbsent, second.Evidence)
-}
-
-func TestCleanupAccount_AbsentKeyIsIdempotent(t *testing.T) {
-	svc, _ := newOverrideService(t)
-	result, err := svc.CleanupAccount(t.Context(), "333333333333")
-	require.NoError(t, err)
-	assert.Equal(t, OverrideCleanupAbsent, result.Disposition)
-	assert.Equal(t, evidenceObservedAbsent, result.Evidence)
-}
-
-func TestCleanupAccount_OtherAccountsUntouched(t *testing.T) {
-	svc, _ := newOverrideService(t)
-	ctx := t.Context()
-	accountA, accountB := testAccount, "222222222222"
-	require.NoError(t, svc.PutAccountQuota(ctx, accountA, Overrides{VCPUs: ptr(32)}, "operator"))
-	require.NoError(t, svc.PutAccountQuota(ctx, accountB, Overrides{VCPUs: ptr(8)}, "operator"))
-
-	result, err := svc.CleanupAccount(ctx, accountA)
-	require.NoError(t, err)
-	assert.Equal(t, OverrideCleanupDeleted, result.Disposition)
-
-	over, _, err := svc.GetAccountQuota(ctx, accountB)
-	require.NoError(t, err)
-	require.NotNil(t, over.VCPUs)
-	assert.Equal(t, 8, *over.VCPUs)
-}
-
-func TestCleanupAccount_RefusesGlobalAccountID(t *testing.T) {
-	svc, _ := newOverrideService(t)
-	result, err := svc.CleanupAccount(t.Context(), awsidentifiers.GlobalAccountID)
-	assert.Error(t, err)
-	assert.Nil(t, result)
-}
-
-func TestCleanupAccount_RefusesEmptyAccountID(t *testing.T) {
-	svc, _ := newOverrideService(t)
-	result, err := svc.CleanupAccount(t.Context(), "")
-	assert.Error(t, err)
-	assert.Nil(t, result)
-}
-
-func TestCleanupAccount_RefusesInvalidAccountID(t *testing.T) {
-	svc, _ := newOverrideService(t)
-	result, err := svc.CleanupAccount(t.Context(), "not-an-account")
-	assert.Error(t, err)
-	assert.Nil(t, result)
-}
-
-func TestCleanupAccount_WithoutBucketFails(t *testing.T) {
-	svc := New(baseLimits(), nil)
-	result, err := svc.CleanupAccount(t.Context(), testAccount)
-	assert.Error(t, err)
-	require.NotNil(t, result)
-	assert.Equal(t, OverrideCleanupFailed, result.Disposition)
-}
-
-// A concurrent Put landing between CleanupAccount's read and its
-// revision-guarded delete must not be lost silently: the delete is rejected
-// and retried against the revision the write left behind.
-func TestCleanupAccount_ConcurrentPutDuringCleanupRetriesAndDeletes(t *testing.T) {
-	svc, _ := newOverrideService(t)
-	ctx := t.Context()
-	require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(32)}, "operator"))
-
-	fake := &overrideCleanupFakeKV{KeyValue: svc.overrides}
-	fake.beforeDelete = func(attempt int) error {
-		if attempt != 1 {
-			return nil
-		}
-		// Simulate a racing override write landing after CleanupAccount's
-		// read but before its guarded delete, through the real KV so the
-		// delete's expected revision is now stale.
-		return svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator")
-	}
-	svc.overrides = fake
-
-	result, err := svc.CleanupAccount(ctx, testAccount)
-	require.NoError(t, err)
-	assert.Equal(t, OverrideCleanupDeleted, result.Disposition)
-	assert.Equal(t, 2, fake.deleteCalls, "the first delete is rejected by the race and retried once")
-}
-
-// A revision that keeps moving under sustained concurrent writes must not
-// spin forever: cleanup gives up and reports failed, bounded by attempts.
-func TestCleanupAccount_PersistentConcurrentWritesFailsBounded(t *testing.T) {
-	svc, _ := newOverrideService(t)
-	ctx := t.Context()
-	require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(32)}, "operator"))
-
-	fake := &overrideCleanupFakeKV{KeyValue: svc.overrides}
-	fake.beforeDelete = func(int) error {
-		return svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator")
-	}
-	svc.overrides = fake
-
-	result, err := svc.CleanupAccount(ctx, testAccount)
-	assert.Error(t, err)
-	require.NotNil(t, result)
-	assert.Equal(t, OverrideCleanupFailed, result.Disposition)
-	assert.Equal(t, cleanupMaxAttempts, fake.deleteCalls)
-}
-
-// A store error must report failed, never absent: absence is a claim about
-// what the store holds, which an error gives no basis to make.
-func TestCleanupAccount_StoreErrorIsFailedNotAbsent(t *testing.T) {
-	svc, _ := newOverrideService(t)
-	fake := &overrideCleanupFakeKV{KeyValue: svc.overrides, getErr: errors.New("kv unavailable")}
-	svc.overrides = fake
-
-	result, err := svc.CleanupAccount(t.Context(), testAccount)
-	assert.Error(t, err)
-	require.NotNil(t, result)
-	assert.Equal(t, OverrideCleanupFailed, result.Disposition)
-}

spinifex/domains/admission/quota/live.go

@@ -39,7 +39,11 @@ func (s *Service) EnforceVPCs(ctx context.Context, natsConn *nats.Conn, accountI
 	if err != nil {
 		return err
 	}
-	return exceeds(len(out.Vpcs), want, s.limitsFor(ctx, accountID).VPCs)
+	limits, err := s.limitsFor(ctx, accountID)
+	if err != nil {
+		return err
+	}
+	return exceeds(len(out.Vpcs), want, limits.VPCs)
 }
 
 // EnforceSubnets gates CreateSubnet on the account's live DescribeSubnets count.
@@ -52,7 +56,11 @@ func (s *Service) EnforceSubnets(ctx context.Context, natsConn *nats.Conn, accou
 	if err != nil {
 		return err
 	}
-	return exceeds(len(out.Subnets), want, s.limitsFor(ctx, accountID).Subnets)
+	limits, err := s.limitsFor(ctx, accountID)
+	if err != nil {
+		return err
+	}
+	return exceeds(len(out.Subnets), want, limits.Subnets)
 }
 
 // EnforceEIPs gates AllocateAddress on the account's live DescribeAddresses count.
@@ -64,7 +72,11 @@ func (s *Service) EnforceEIPs(ctx context.Context, natsConn *nats.Conn, accountI
 	if err != nil {
 		return err
 	}
-	return exceeds(len(out.Addresses), want, s.limitsFor(ctx, accountID).EIPs)
+	limits, err := s.limitsFor(ctx, accountID)
+	if err != nil {
+		return err
+	}
+	return exceeds(len(out.Addresses), want, limits.EIPs)
 }
 
 // EnforceLoadBalancers gates CreateLoadBalancer on the account's live count.
@@ -77,7 +89,11 @@ func (s *Service) EnforceLoadBalancers(ctx context.Context, natsConn *nats.Conn,
 	if err != nil {
 		return err
 	}
-	return exceeds(len(out.LoadBalancers), want, s.limitsFor(ctx, accountID).LoadBalancers)
+	limits, err := s.limitsFor(ctx, accountID)
+	if err != nil {
+		return err
+	}
+	return exceeds(len(out.LoadBalancers), want, limits.LoadBalancers)
 }
 
 // EnforceRDSInstances gates CreateDBInstance on the account's live count. A DB
@@ -91,5 +107,9 @@ func (s *Service) EnforceRDSInstances(ctx context.Context, natsConn *nats.Conn,
 	if err != nil {
 		return err
 	}
-	return exceeds(len(out.DBInstances), want, s.limitsFor(ctx, accountID).RDSInstances)
+	limits, err := s.limitsFor(ctx, accountID)
+	if err != nil {
+		return err
+	}
+	return exceeds(len(out.DBInstances), want, limits.RDSInstances)
 }

spinifex/domains/admission/quota/override.go

@@ -8,12 +8,16 @@ import (
 	"log/slog"
 	"time"
 
+	awserrors "github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
+	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvutil"
 	"github.com/nats-io/nats.go/jetstream"
 )
 
 // Overrides raises or lowers individual dimensions for one account. Every field
 // is a pointer so an explicit 0 (a limit that denies everything) is distinct
-// from an absent field, which inherits the configured value.
+// from an absent field, which inherits the configured value. This is also the
+// admin wire type: an admin request body can carry only these fields, so it can
+// never set or clear a stored record's state.
 type Overrides struct {
 	VCPUs         *int `json:"vcpus,omitempty"`
 	VPCs          *int `json:"vpcs,omitempty"`
@@ -29,11 +33,62 @@ type Overrides struct {
 	UpdatedBy string `json:"updated_by,omitempty"`
 }
 
-// Empty reports whether no dimension is overridden, in which case the record
-// carries nothing but provenance and the account resolves to the config.
-func (o Overrides) Empty() bool {
-	return o.VCPUs == nil && o.VPCs == nil && o.Subnets == nil && o.EIPs == nil &&
-		o.VolumesGiB == nil && o.Volumes == nil && o.RDSInstances == nil && o.LoadBalancers == nil
+// OverrideState is a stored override record's lifecycle state. "" is active;
+// a record written before the state field existed has none and stays active.
+type OverrideState string
+
+// OverrideStateSealed marks a record no writer may change or recreate.
+const OverrideStateSealed OverrideState = "sealed"
+
+// overrideRecord is the stored form: the admin-settable fields inline, plus a
+// state no admin request can read or write.
+type overrideRecord struct {
+	Overrides
+
+	State OverrideState `json:"state,omitempty"`
+}
+
+// errOverridesSealed marks a read or write that found a sealed record, so a
+// caller maps it to Blocked without logging it as a fault.
+var errOverridesSealed = errors.New(awserrors.ErrorBlocked)
+
+// errOverridesContended marks a write whose CAS retries were exhausted under
+// concurrent writers, bounding the retry loop instead of spinning.
+var errOverridesContended = errors.New("quota override: CAS retries exhausted")
+
+// overrideMaxAttempts bounds a write's retries against a revision that keeps
+// moving under concurrent writers, so it fails closed instead of spinning.
+const overrideMaxAttempts = 5
+
+// checkActive rejects a sealed or unknown-state record; "" is active and
+// passes through. A caller must never read or mutate a record that failed
+// this as if it were active.
+func checkActive(rec overrideRecord) error {
+	switch rec.State {
+	case "":
+		return nil
+	case OverrideStateSealed:
+		return errOverridesSealed
+	default:
+		return fmt.Errorf("quota override: unknown state %q", rec.State)
+	}
+}
+
+// mapOverrideError turns a recordFor, PutAccountQuota or clear failure into
+// the AWS error a caller returns: sealed is an expected, unlogged Blocked;
+// everything else is InternalError, logged with the account ID.
+func mapOverrideError(ctx context.Context, accountID string, err error) error {
+	switch {
+	case err == nil:
+		return nil
+	case errors.Is(err, errOverridesSealed):
+		return errors.New(awserrors.ErrorBlocked)
+	case errors.Is(err, errOverridesContended):
+		slog.ErrorContext(ctx, "Quota override CAS retries exhausted under contention", "accountID", accountID)
+	default:
+		slog.ErrorContext(ctx, "Quota override operation failed", "accountID", accountID, "err", err)
+	}
+	return errors.New(awserrors.ErrorInternalError)
 }
 
 // apply overlays the set fields onto base, leaving Enabled and the Bedrock
@@ -60,72 +115,128 @@ func (o Overrides) apply(base Limits) Limits {
 	return out
 }
 
-// limitsFor resolves the effective limits for accountID: its overrides where
-// set, the configured limits everywhere else. A read failure falls back to the
-// configured limits rather than to no limit, so a KV outage cannot lift a cap.
-func (s *Service) limitsFor(ctx context.Context, accountID string) Limits {
-	over, err := s.overridesFor(ctx, accountID)
-	if err != nil {
-		slog.WarnContext(ctx, "quota: override read failed, using configured limits",
-			"account", accountID, "err", err)
-		return s.limits
-	}
-	return over.apply(s.limits)
+// Empty reports whether no dimension is overridden, in which case the record
+// carries nothing but provenance and the account resolves to the config.
+func (o Overrides) Empty() bool {
+	return o.VCPUs == nil && o.VPCs == nil && o.Subnets == nil && o.EIPs == nil &&
+		o.VolumesGiB == nil && o.Volumes == nil && o.RDSInstances == nil && o.LoadBalancers == nil
 }
 
-// overridesFor reads accountID's override record. An absent key is the normal
-// case for nearly every account and yields the zero value, not an error; only a
-// genuine read or decode failure is returned.
-func (s *Service) overridesFor(ctx context.Context, accountID string) (Overrides, error) {
+// recordFor reads and decodes accountID's stored override record. An absent
+// key is the active zero value, not an error; only a genuine read or decode
+// failure is returned.
+func (s *Service) recordFor(ctx context.Context, accountID string) (overrideRecord, error) {
 	if s == nil || s.overrides == nil {
-		return Overrides{}, nil
+		return overrideRecord{}, nil
 	}
 	entry, err := s.overrides.Get(ctx, accountID)
 	if errors.Is(err, jetstream.ErrKeyNotFound) {
-		return Overrides{}, nil
+		return overrideRecord{}, nil
 	}
 	if err != nil {
-		return Overrides{}, fmt.Errorf("read quota override for %s: %w", accountID, err)
+		return overrideRecord{}, fmt.Errorf("read quota override for %s: %w", accountID, err)
 	}
-	var over Overrides
-	if err := json.Unmarshal(entry.Value(), &over); err != nil {
-		return Overrides{}, fmt.Errorf("decode quota override for %s: %w", accountID, err)
+	var rec overrideRecord
+	if err := json.Unmarshal(entry.Value(), &rec); err != nil {
+		return overrideRecord{}, fmt.Errorf("decode quota override for %s: %w", accountID, err)
 	}
-	return over, nil
+	return rec, nil
 }
 
-// GetAccountQuota returns accountID's stored overrides and the limits they
-// resolve to. A read failure is surfaced here rather than masked, because an
-// operator inspecting a quota must not be shown the fallback as if it were real.
+// limitsFor resolves accountID's effective limits: overrides where set, the
+// config elsewhere. A read or decode failure falls back to the config so an
+// outage cannot lift a cap; a sealed or unknown-state record refuses instead.
+func (s *Service) limitsFor(ctx context.Context, accountID string) (Limits, error) {
+	rec, err := s.recordFor(ctx, accountID)
+	if err != nil {
+		slog.WarnContext(ctx, "quota: override read failed, using configured limits",
+			"account", accountID, "err", err)
+		return s.limits, nil
+	}
+	if err := checkActive(rec); err != nil {
+		return Limits{}, mapOverrideError(ctx, accountID, err)
+	}
+	return rec.apply(s.limits), nil
+}
+
+// GetAccountQuota returns accountID's stored overrides and resolved limits.
+// Read failures surface rather than show the fallback as real; a sealed or
+// unknown-state record returns Blocked or InternalError.
 func (s *Service) GetAccountQuota(ctx context.Context, accountID string) (Overrides, Limits, error) {
-	over, err := s.overridesFor(ctx, accountID)
+	rec, err := s.recordFor(ctx, accountID)
 	if err != nil {
 		return Overrides{}, Limits{}, err
 	}
-	return over, over.apply(s.limits), nil
+	if err := checkActive(rec); err != nil {
+		return Overrides{}, Limits{}, mapOverrideError(ctx, accountID, err)
+	}
+	return rec.Overrides, rec.apply(s.limits), nil
 }
 
-// PutAccountQuota stores accountID's overrides, stamping provenance. An empty
-// set deletes the record so the account resolves to the config, which keeps
-// "has this account been touched?" a single existence check.
+// PutAccountQuota stores or clears accountID's overrides, stamping provenance
+// on a store. Both branches are revision-guarded and refuse a sealed record,
+// so a seal landing between a read and a write fails the CAS.
 func (s *Service) PutAccountQuota(ctx context.Context, accountID string, over Overrides, updatedBy string) error {
 	if s == nil || s.overrides == nil {
 		return errors.New("quota: override bucket not configured")
 	}
 	if over.Empty() {
-		if err := s.overrides.Delete(ctx, accountID); err != nil && !errors.Is(err, jetstream.ErrKeyNotFound) {
-			return fmt.Errorf("clear quota override for %s: %w", accountID, err)
-		}
-		return nil
+		return s.clearAccountQuota(ctx, accountID)
 	}
+
 	over.UpdatedAt = time.Now().UTC().Format(time.RFC3339)
 	over.UpdatedBy = updatedBy
-	data, err := json.Marshal(over)
-	if err != nil {
-		return fmt.Errorf("encode quota override for %s: %w", accountID, err)
-	}
-	if _, err := s.overrides.Put(ctx, accountID, data); err != nil {
-		return fmt.Errorf("store quota override for %s: %w", accountID, err)
-	}
-	return nil
+
+	_, err := kvutil.Update(ctx, s.overrides, accountID, kvutil.CASConfig{
+		CreateIfAbsent: true,
+		Attempts:       overrideMaxAttempts,
+		Exhausted:      func(string, int) error { return errOverridesContended },
+	}, func(rec *overrideRecord) (bool, error) {
+		if err := checkActive(*rec); err != nil {
+			return false, err
+		}
+		rec.Overrides = over
+		// Every success is a leader-accepted write over an active value: a
+		// seal landing between the read and the write fails this CAS.
+		return true, nil
+	})
+	return mapOverrideError(ctx, accountID, err)
+}
+
+// clearAccountQuota deletes an active override at the revision it read. An
+// observed-absent key writes nothing, so a stale read cannot unseal; a sealed
+// record refuses rather than being erased and later recreated active.
+func (s *Service) clearAccountQuota(ctx context.Context, accountID string) error {
+	for range overrideMaxAttempts {
+		entry, err := s.overrides.Get(ctx, accountID)
+		if errors.Is(err, jetstream.ErrKeyNotFound) {
+			return nil
+		}
+		if err != nil {
+			return mapOverrideError(ctx, accountID, fmt.Errorf("read quota override for clear: %w", err))
+		}
+
+		var rec overrideRecord
+		if err := json.Unmarshal(entry.Value(), &rec); err != nil {
+			return mapOverrideError(ctx, accountID, fmt.Errorf("decode quota override for clear: %w", err))
+		}
+		if err := checkActive(rec); err != nil {
+			return mapOverrideError(ctx, accountID, err)
+		}
+
+		err = s.overrides.Delete(ctx, accountID, jetstream.LastRevision(entry.Revision()))
+		switch {
+		case err == nil:
+			return nil
+		case errors.Is(err, jetstream.ErrKeyNotFound):
+			return nil
+		case isRevisionConflict(err):
+			// A concurrent write landed between the read and the guarded
+			// delete; retry against whatever revision now holds.
+			continue
+		default:
+			return mapOverrideError(ctx, accountID, fmt.Errorf("clear quota override: %w", err))
+		}
+	}
+	return mapOverrideError(ctx, accountID, errOverridesContended)
 }

spinifex/domains/admission/quota/override_kv_test.go

@@ -31,7 +31,8 @@ func newOverrideService(t *testing.T) (*Service, jetstream.KeyValue) {
 func TestLimitsForAbsentKeyInherits(t *testing.T) {
 	svc, _ := newOverrideService(t)
 
-	got := svc.limitsFor(t.Context(), "123456789012")
+	got, err := svc.limitsFor(t.Context(), "123456789012")
+	require.NoError(t, err)
 	require.Equal(t, baseLimits(), got)
 }
 
@@ -44,10 +45,14 @@ func TestLimitsForIsPerAccount(t *testing.T) {
 	require.NoError(t, svc.PutAccountQuota(t.Context(), upgraded,
 		Overrides{VCPUs: ptr(32), RDSInstances: ptr(8)}, "operator"))
 
-	require.Equal(t, 32, svc.limitsFor(t.Context(), upgraded).VCPUs)
-	require.Equal(t, 8, svc.limitsFor(t.Context(), upgraded).RDSInstances)
-	require.Equal(t, baseLimits().VCPUs, svc.limitsFor(t.Context(), sandbox).VCPUs)
-	require.Equal(t, baseLimits().RDSInstances, svc.limitsFor(t.Context(), sandbox).RDSInstances)
+	upgradedLimits, err := svc.limitsFor(t.Context(), upgraded)
+	require.NoError(t, err)
+	sandboxLimits, err := svc.limitsFor(t.Context(), sandbox)
+	require.NoError(t, err)
+	require.Equal(t, 32, upgradedLimits.VCPUs)
+	require.Equal(t, 8, upgradedLimits.RDSInstances)
+	require.Equal(t, baseLimits().VCPUs, sandboxLimits.VCPUs)
+	require.Equal(t, baseLimits().RDSInstances, sandboxLimits.RDSInstances)
 }
 
 // A stored override must round-trip through KV with its provenance, and an
@@ -79,7 +84,9 @@ func TestPutAccountQuotaEmptyClears(t *testing.T) {
 
 	_, err := kv.Get(t.Context(), account)
 	require.ErrorIs(t, err, jetstream.ErrKeyNotFound)
-	require.Equal(t, baseLimits(), svc.limitsFor(t.Context(), account))
+	limits, err := svc.limitsFor(t.Context(), account)
+	require.NoError(t, err)
+	require.Equal(t, baseLimits(), limits)
 }
 
 // Clearing an account that never had a record must succeed rather than fail on
@@ -97,8 +104,10 @@ func TestPutAccountQuotaExplicitZero(t *testing.T) {
 
 	require.NoError(t, svc.PutAccountQuota(t.Context(), account, Overrides{VPCs: ptr(0)}, "operator"))
 
-	require.Equal(t, 0, svc.limitsFor(t.Context(), account).VPCs)
-	require.Error(t, exceeds(0, 1, svc.limitsFor(t.Context(), account).VPCs))
+	limits, err := svc.limitsFor(t.Context(), account)
+	require.NoError(t, err)
+	require.Equal(t, 0, limits.VPCs)
+	require.Error(t, exceeds(0, 1, limits.VPCs))
 }
 
 // A corrupt record is a read failure, not an absent one: it must fall back to
@@ -110,7 +119,9 @@ func TestLimitsForCorruptRecordFailsClosed(t *testing.T) {
 	_, err := kv.Put(t.Context(), account, []byte("{not json"))
 	require.NoError(t, err)
 
-	require.Equal(t, baseLimits(), svc.limitsFor(t.Context(), account),
+	limits, err := svc.limitsFor(t.Context(), account)
+	require.NoError(t, err, "a decode failure must fall back, not error, from limitsFor")
+	require.Equal(t, baseLimits(), limits,
 		"a corrupt override must fall back to the configured limits")
 
 	_, _, err = svc.GetAccountQuota(t.Context(), account)

spinifex/domains/admission/quota/override_test.go

@@ -126,7 +126,11 @@ func TestOverridesEmpty(t *testing.T) {
 // fall through to the configured limits rather than panicking.
 func TestLimitsForWithoutBucket(t *testing.T) {
 	svc := New(baseLimits(), nil)
-	if got := svc.limitsFor(t.Context(), "123456789012"); got != baseLimits() {
+	got, err := svc.limitsFor(t.Context(), "123456789012")
+	if err != nil {
+		t.Fatalf("limitsFor error = %v, want nil", err)
+	}
+	if got != baseLimits() {
 		t.Fatalf("limitsFor = %+v, want the configured limits", got)
 	}
 }

spinifex/domains/admission/quota/override_writes_guarded_test.go

@@ -0,0 +1,75 @@
+package quota
+
+//test:in-package — pins what AccountQuotaOverrideSealCapabilityVersion 1 will
+// assert: nothing in this package writes the override bucket except through
+// PutAccountQuota's CAS loop, clearAccountQuota's guarded delete or
+// SealAccountOverrides' CAS loop.
+
+import (
+	"go/ast"
+	"go/parser"
+	"go/token"
+	"path/filepath"
+	"runtime"
+	"testing"
+)
+
+// overrideWriteFiles are the only files in this package that touch the
+// override bucket. The package also holds the vCPU usage and Bedrock usage
+// buckets, which write through their own stores and are out of scope here;
+// scanning the whole package directory, as the EC2 settings equivalent does,
+// would false-flag those unrelated writes.
+var overrideWriteFiles = []string{"override.go", "seal.go"}
+
+func quotaPackageDir(t *testing.T) string {
+	t.Helper()
+	_, file, _, ok := runtime.Caller(0)
+	if !ok {
+		t.Fatal("cannot locate test source")
+	}
+	return filepath.Dir(file)
+}
+
+// No file owning the override bucket may call Put, PutString, or Delete
+// without a revision option: every write must go through PutAccountQuota's
+// CAS loop, clearAccountQuota's guarded delete, or SealAccountOverrides' CAS
+// loop.
+func TestOverrideWritesAreRevisionGuarded(t *testing.T) {
+	root := quotaPackageDir(t)
+	fset := token.NewFileSet()
+	scanned := 0
+
+	for _, name := range overrideWriteFiles {
+		path := filepath.Join(root, name)
+		f, err := parser.ParseFile(fset, path, nil, 0)
+		if err != nil {
+			t.Fatalf("parse %s: %v", path, err)
+		}
+		scanned++
+
+		ast.Inspect(f, func(n ast.Node) bool {
+			call, ok := n.(*ast.CallExpr)
+			if !ok {
+				return true
+			}
+			sel, ok := call.Fun.(*ast.SelectorExpr)
+			if !ok {
+				return true
+			}
+			switch sel.Sel.Name {
+			case "Put", "PutString":
+				t.Errorf("%s: unconditional %s call bypasses the revision-guarded override write; use PutAccountQuota or SealAccountOverrides instead",
+					fset.Position(call.Pos()), sel.Sel.Name)
+			case "Delete":
+				if len(call.Args) < 3 {
+					t.Errorf("%s: Delete call with no revision option bypasses the guarded clear; pin it to the read revision",
+						fset.Position(call.Pos()))
+				}
+			}
+			return true
+		})
+	}
+	if scanned == 0 {
+		t.Fatalf("scanned no override-write files under %s; the invariant would pass vacuously", root)
+	}
+}

spinifex/domains/admission/quota/quota.go

@@ -90,6 +90,14 @@ type Service struct {
 	// overrides holds the per-account limit overrides. Nil leaves every account
 	// on the configured limits, which is also what an absent key means.
 	overrides jetstream.KeyValue
+
+	// js opens the override bucket for SealAccountOverrides when quotas were
+	// disabled at startup; nil makes such a seal refuse rather than create it.
+	js jetstream.KeyValueManager
+
+	// terminationChecker is the seal precondition; SealAccountOverrides
+	// refuses until it is wired.
+	terminationChecker AccountTerminationChecker
 }
 
 // New constructs a quota Service from the configured limits and the gateway-owned
@@ -125,6 +133,16 @@ func (s *Service) SetOverrides(kv jetstream.KeyValue) {
 	s.overrides = kv
 }
 
+// SetJetStream wires the client SealAccountOverrides uses to get-or-create
+// the override bucket when this gateway disabled quotas. Called once at
+// wiring time, regardless of whether quotas are enabled here.
+func (s *Service) SetJetStream(js jetstream.KeyValueManager) {
+	if s == nil {
+		return
+	}
+	s.js = js
+}
+
 // Exempt returns true for the global/system account and whenever quotas are
 // disabled, so those callers bypass every quota check. A nil Service (quotas
 // never built) is treated as disabled so callers need no nil guard.

spinifex/domains/admission/quota/seal.go

@@ -0,0 +1,163 @@
+package quota
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"log/slog"
+
+	awserrors "github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
+	awsidentifiers "github.com/mulgadc/spinifex/spinifex/foundation/aws/identifiers"
+	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvutil"
+	"github.com/nats-io/nats.go/jetstream"
+)
+
+// sealMaxAttempts bounds SealAccountOverrides' retry loop against a revision
+// that keeps moving under concurrent writes, so it fails closed instead of
+// spinning.
+const sealMaxAttempts = 5
+
+// sealedRecord is the fixed value a seal writes. No dimension or provenance
+// survives, so no tenant commercial setting outlives the seal.
+var sealedRecord = &overrideRecord{State: OverrideStateSealed}
+
+// OverrideSeal is SealAccountOverrides' result: the revision at which the
+// stream leader accepted the sealed value.
+type OverrideSeal struct {
+	Revision uint64
+}
+
+// AccountTerminationChecker reports whether IAM holds accountID as
+// TERMINATING. A missing record is (false, nil); only a failed read is an
+// error.
+type AccountTerminationChecker interface {
+	AccountTerminating(ctx context.Context, accountID string) (bool, error)
+}
+
+// SetAccountTerminationChecker wires the seal precondition; until it is set,
+// SealAccountOverrides refuses.
+func (s *Service) SetAccountTerminationChecker(c AccountTerminationChecker) {
+	if s == nil {
+		return
+	}
+	s.terminationChecker = c
+}
+
+// checkTerminating runs before any override-KV call, so a refusal never
+// writes or opens the bucket. Not TERMINATING (incl. missing) is
+// IncorrectState; a read failure or unwired checker is InternalError.
+func (s *Service) checkTerminating(ctx context.Context, accountID string) error {
+	if s.terminationChecker == nil {
+		slog.ErrorContext(ctx, "Quota override seal has no termination checker wired", "accountID", accountID)
+		return errors.New(awserrors.ErrorInternalError)
+	}
+	terminating, err := s.terminationChecker.AccountTerminating(ctx, accountID)
+	if err != nil {
+		slog.ErrorContext(ctx, "Failed to read account status for seal precondition", "accountID", accountID, "err", err)
+		return errors.New(awserrors.ErrorInternalError)
+	}
+	if !terminating {
+		return awserrors.Errorf(awserrors.ErrorIncorrectState, "account %s is not TERMINATING", accountID)
+	}
+	return nil
+}
+
+// isRevisionConflict reports a lost CAS race to retry, including the bare
+// wrong-last-sequence APIError from Create's retry against a delete marker.
+// A private copy of domains/ec2/account's helper, which this package may not import.
+func isRevisionConflict(err error) bool {
+	if errors.Is(err, jetstream.ErrKeyExists) || errors.Is(err, jetstream.ErrKeyRevisionMismatch) {
+		return true
+	}
+	if apiErr, ok := errors.AsType[*jetstream.APIError](err); ok {
+		return apiErr.ErrorCode == jetstream.JSErrCodeStreamWrongLastSequence ||
+			apiErr.ErrorCode == jetstream.JSErrCodeStreamWrongLastSequenceConstant
+	}
+	return false
+}
+
+// sealBucket returns this gateway's override bucket, or get-or-creates it
+// when quotas are disabled here. Called only after checkTerminating accepts,
+// so a refused seal never creates the bucket.
+func (s *Service) sealBucket(ctx context.Context) (jetstream.KeyValue, error) {
+	if s.overrides != nil {
+		return s.overrides, nil
+	}
+	if s.js == nil {
+		slog.ErrorContext(ctx, "Quota override seal has no JetStream client wired to open the override bucket")
+		return nil, errors.New(awserrors.ErrorInternalError)
+	}
+	kv, err := kvutil.GetOrCreateBucket(ctx, s.js, KVBucketAccountQuota, 1)
+	if err != nil {
+		slog.ErrorContext(ctx, "Failed to open quota override bucket for seal", "err", err)
+		return nil, errors.New(awserrors.ErrorInternalError)
+	}
+	return kv, nil
+}
+
+// SealAccountOverrides seals accountID's override record so no later writer
+// can recreate it. Success is only a leader-accepted write, evidenced by the
+// revision; a missing key is created sealed, never reported from absence.
+func (s *Service) SealAccountOverrides(ctx context.Context, accountID string) (*OverrideSeal, error) {
+	if accountID == "" {
+		return nil, awserrors.Errorf(awserrors.ErrorMissingParameter, "accountID is required to seal quota overrides")
+	}
+	if !awsidentifiers.IsAccountID(accountID) {
+		return nil, awserrors.Errorf(awserrors.ErrorInvalidParameterValue, "accountID %q is not a valid AWS account ID", accountID)
+	}
+	if accountID == awsidentifiers.GlobalAccountID {
+		return nil, awserrors.Errorf(awserrors.ErrorInvalidParameterValue, "quota override seal must not target the global account")
+	}
+
+	if err := s.checkTerminating(ctx, accountID); err != nil {
+		return nil, err
+	}
+
+	overrides, err := s.sealBucket(ctx)
+	if err != nil {
+		return nil, err
+	}
+
+	data, err := json.Marshal(sealedRecord)
+	if err != nil {
+		slog.ErrorContext(ctx, "Failed to encode sealed quota override record", "accountID", accountID, "err", err)
+		return nil, errors.New(awserrors.ErrorInternalError)
+	}
+
+	// accountID is never resolved through any fallback here: a seal must act
+	// on exactly accountID and never fall back to the global key.
+	for range sealMaxAttempts {
+		var rev uint64
+		entry, getErr := overrides.Get(ctx, accountID)
+		switch {
+		case getErr == nil:
+			rev = entry.Revision()
+		case errors.Is(getErr, jetstream.ErrKeyNotFound):
+			rev = 0
+		default:
+			slog.ErrorContext(ctx, "Failed to read quota override for seal", "accountID", accountID, "err", getErr)
+			return nil, errors.New(awserrors.ErrorInternalError)
+		}
+
+		var newRev uint64
+		var writeErr error
+		if rev == 0 {
+			newRev, writeErr = overrides.Create(ctx, accountID, data)
+		} else {
+			newRev, writeErr = overrides.Update(ctx, accountID, data, rev)
+		}
+
+		switch {
+		case writeErr == nil:
+			return &OverrideSeal{Revision: newRev}, nil
+		case isRevisionConflict(writeErr):
+			continue
+		default:
+			slog.ErrorContext(ctx, "Failed to write sealed quota override record", "accountID", accountID, "err", writeErr)
+			return nil, errors.New(awserrors.ErrorInternalError)
+		}
+	}
+
+	slog.ErrorContext(ctx, "Quota override seal exhausted attempts against concurrent writes", "accountID", accountID, "attempts", sealMaxAttempts)
+	return nil, errors.New(awserrors.ErrorInternalError)
+}

spinifex/domains/admission/quota/seal_test.go

@@ -0,0 +1,806 @@
+package quota
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"sync"
+	"testing"
+
+	"github.com/mulgadc/spinifex/internal/testkit"
+	awserrors "github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
+	awsidentifiers "github.com/mulgadc/spinifex/spinifex/foundation/aws/identifiers"
+	"github.com/nats-io/nats.go/jetstream"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+// stubTerminationChecker answers AccountTerminating from a fixed set, for
+// tests that exercise the KV paths rather than IAM itself.
+type stubTerminationChecker struct {
+	terminating map[string]bool
+	err         error
+}
+
+func (c *stubTerminationChecker) AccountTerminating(_ context.Context, accountID string) (bool, error) {
+	if c.err != nil {
+		return false, c.err
+	}
+	return c.terminating[accountID], nil
+}
+
+// terminatingChecker answers true for exactly accountID.
+func terminatingChecker(accountID string) *stubTerminationChecker {
+	return &stubTerminationChecker{terminating: map[string]bool{accountID: true}}
+}
+
+// newSealService builds a Service over a real JetStream KV bucket with the
+// override bucket already attached, mirroring a quota-enabled gateway.
+func newSealService(t *testing.T) (*Service, jetstream.JetStream) {
+	t.Helper()
+	_, _, js := testutil.StartTestJetStream(t)
+	kv, err := js.CreateKeyValue(t.Context(), jetstream.KeyValueConfig{
+		Bucket:  KVBucketAccountQuota,
+		History: 1,
+	})
+	require.NoError(t, err)
+	usage, err := js.CreateKeyValue(t.Context(), jetstream.KeyValueConfig{
+		Bucket:  KVBucketAccountUsage,
+		History: 1,
+	})
+	require.NoError(t, err)
+
+	svc := New(baseLimits(), usage)
+	svc.SetOverrides(kv)
+	svc.SetJetStream(js)
+	return svc, js
+}
+
+// newDisabledSealService builds a Service with only the JetStream client
+// wired, mirroring a quota-disabled gateway: the override bucket does not
+// exist until a seal creates it.
+func newDisabledSealService(t *testing.T) (*Service, jetstream.JetStream) {
+	t.Helper()
+	_, _, js := testutil.StartTestJetStream(t)
+	svc := New(baseLimits(), nil)
+	svc.SetJetStream(js)
+	return svc, js
+}
+
+// lastOverrideSeq reports the override stream's last sequence, so a test can
+// assert a failed attempt wrote nothing.
+func lastOverrideSeq(t *testing.T, js jetstream.JetStream) uint64 {
+	t.Helper()
+	stream, err := js.Stream(context.Background(), "KV_"+KVBucketAccountQuota)
+	require.NoError(t, err)
+	info, err := stream.Info(context.Background())
+	require.NoError(t, err)
+	return info.State.LastSeq
+}
+
+// mutationCase is one of PutAccountQuota's two mutations, called uniformly so
+// a test can table over them.
+type mutationCase struct {
+	name string
+	call func(svc *Service, ctx context.Context) error
+}
+
+func mutationTable() []mutationCase {
+	return []mutationCase{
+		{"Set", func(svc *Service, ctx context.Context) error {
+			return svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(32)}, "operator")
+		}},
+		{"Clear", func(svc *Service, ctx context.Context) error {
+			return svc.PutAccountQuota(ctx, testAccount, Overrides{}, "operator")
+		}},
+	}
+}
+
+// seedOverrideState prepares testAccount's record for a race test: present
+// (an active record), absent (never written) or tombstone (written then
+// deleted, leaving a delete marker Get hides).
+func seedOverrideState(t *testing.T, svc *Service, state string) {
+	t.Helper()
+	ctx := context.Background()
+	switch state {
+	case "present":
+		require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator"))
+	case "tombstone":
+		require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator"))
+		require.NoError(t, svc.overrides.Delete(ctx, testAccount))
+	case "absent":
+	default:
+		t.Fatalf("unknown seed state %q", state)
+	}
+}
+
+// sealRaceKV wraps a real KeyValue so a test can inject a deterministic race
+// between a caller's read and its guarded write (Create, Update or Delete),
+// instead of relying on goroutine timing.
+type sealRaceKV struct {
+	jetstream.KeyValue
+
+	writeCalls  int
+	beforeWrite func(attempt int) error
+}
+
+func (f *sealRaceKV) Create(ctx context.Context, key string, value []byte, opts ...jetstream.KVCreateOpt) (uint64, error) {
+	f.writeCalls++
+	if f.beforeWrite != nil {
+		if err := f.beforeWrite(f.writeCalls); err != nil {
+			return 0, err
+		}
+	}
+	return f.KeyValue.Create(ctx, key, value, opts...)
+}
+
+func (f *sealRaceKV) Update(ctx context.Context, key string, value []byte, revision uint64) (uint64, error) {
+	f.writeCalls++
+	if f.beforeWrite != nil {
+		if err := f.beforeWrite(f.writeCalls); err != nil {
+			return 0, err
+		}
+	}
+	return f.KeyValue.Update(ctx, key, value, revision)
+}
+
+func (f *sealRaceKV) Delete(ctx context.Context, key string, opts ...jetstream.KVDeleteOpt) error {
+	f.writeCalls++
+	if f.beforeWrite != nil {
+		if err := f.beforeWrite(f.writeCalls); err != nil {
+			return err
+		}
+	}
+	return f.KeyValue.Delete(ctx, key, opts...)
+}
+
+// staleGetKV always answers Get with a fixed snapshot taken at construction,
+// so a caller's CAS keeps retrying against a revision that can never win.
+type staleGetKV struct {
+	jetstream.KeyValue
+
+	entry jetstream.KeyValueEntry
+}
+
+func (f *staleGetKV) Get(context.Context, string) (jetstream.KeyValueEntry, error) {
+	return f.entry, nil
+}
+
+// staleNotFoundKV always answers Get as absent, regardless of what the real
+// bucket holds.
+type staleNotFoundKV struct {
+	jetstream.KeyValue
+}
+
+func (f *staleNotFoundKV) Get(context.Context, string) (jetstream.KeyValueEntry, error) {
+	return nil, jetstream.ErrKeyNotFound
+}
+
+// errGetKV always fails Get with a fixed store error.
+type errGetKV struct {
+	jetstream.KeyValue
+
+	err error
+}
+
+func (f *errGetKV) Get(context.Context, string) (jetstream.KeyValueEntry, error) {
+	return nil, f.err
+}
+
+func decodeOverride(t *testing.T, entry jetstream.KeyValueEntry) overrideRecord {
+	t.Helper()
+	var rec overrideRecord
+	require.NoError(t, json.Unmarshal(entry.Value(), &rec))
+	return rec
+}
+
+// A seal landing between a mutation's read and its guarded write is not
+// outlived: the mutation's retry sees the seal and is rejected, and nothing
+// it wrote survives. Clear on an absent or tombstoned key never reaches a
+// write at all (see TestClearOnMissingKeyWritesNothing), so only the four
+// cells with a real write are exercised here.
+func TestSealRaceMutation_SealLandsBetweenReadAndWrite(t *testing.T) {
+	cases := []struct {
+		name  string
+		state string
+		call  func(svc *Service, ctx context.Context) error
+	}{
+		{"Set_present", "present", mutationTable()[0].call},
+		{"Set_absent", "absent", mutationTable()[0].call},
+		{"Set_tombstone", "tombstone", mutationTable()[0].call},
+		{"Clear_present", "present", mutationTable()[1].call},
+	}
+	for _, tc := range cases {
+		t.Run(tc.name, func(t *testing.T) {
+			svc, _ := newSealService(t)
+			ctx := context.Background()
+			seedOverrideState(t, svc, tc.state)
+
+			realKV := svc.overrides
+			sealer := New(baseLimits(), nil)
+			sealer.SetOverrides(realKV)
+			sealer.SetAccountTerminationChecker(terminatingChecker(testAccount))
+
+			var sealed *OverrideSeal
+			race := &sealRaceKV{KeyValue: realKV}
+			race.beforeWrite = func(attempt int) error {
+				if attempt != 1 {
+					return nil
+				}
+				var sealErr error
+				sealed, sealErr = sealer.SealAccountOverrides(ctx, testAccount)
+				return sealErr
+			}
+			svc.overrides = race
+
+			err := tc.call(svc, ctx)
+			require.Error(t, err)
+			assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorBlocked), "got %v", err)
+			require.NotNil(t, sealed)
+
+			entry, err := realKV.Get(ctx, testAccount)
+			require.NoError(t, err)
+			assert.Equal(t, OverrideStateSealed, decodeOverride(t, entry).State)
+			assert.Equal(t, sealed.Revision, entry.Revision(), "no write landed after the seal")
+		})
+	}
+}
+
+// A mutation landing inside a seal's read-write window does not survive it:
+// the seal retries against the mutation's write and wins, leaving the sealed
+// value with no overridden dimension.
+func TestSealRaceMutation_MutationLandsBetweenSealReadAndWrite(t *testing.T) {
+	for _, m := range mutationTable() {
+		for _, state := range []string{"present", "absent", "tombstone"} {
+			t.Run(m.name+"_"+state, func(t *testing.T) {
+				svc, _ := newSealService(t)
+				ctx := context.Background()
+				seedOverrideState(t, svc, state)
+
+				realKV := svc.overrides
+				race := &sealRaceKV{KeyValue: realKV}
+				race.beforeWrite = func(attempt int) error {
+					if attempt != 1 {
+						return nil
+					}
+					_ = m.call(svc, ctx)
+					return nil
+				}
+
+				sealer := New(baseLimits(), nil)
+				sealer.SetOverrides(race)
+				sealer.SetAccountTerminationChecker(terminatingChecker(testAccount))
+
+				seal, err := sealer.SealAccountOverrides(ctx, testAccount)
+				require.NoError(t, err)
+
+				entry, err := realKV.Get(ctx, testAccount)
+				require.NoError(t, err)
+				rec := decodeOverride(t, entry)
+				assert.Equal(t, OverrideStateSealed, rec.State)
+				assert.Nil(t, rec.VCPUs)
+				assert.Equal(t, seal.Revision, entry.Revision())
+			})
+		}
+	}
+}
+
+// Neither mutation can act on an already-sealed record.
+func TestMutationOnSealedRecordIsRejected(t *testing.T) {
+	for _, m := range mutationTable() {
+		t.Run(m.name, func(t *testing.T) {
+			svc, _ := newSealService(t)
+			ctx := context.Background()
+			svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+			_, err := svc.SealAccountOverrides(ctx, testAccount)
+			require.NoError(t, err)
+
+			before, err := svc.overrides.Get(ctx, testAccount)
+			require.NoError(t, err)
+
+			err = m.call(svc, ctx)
+			require.Error(t, err)
+			assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorBlocked), "got %v", err)
+
+			after, err := svc.overrides.Get(ctx, testAccount)
+			require.NoError(t, err)
+			assert.Equal(t, before.Revision(), after.Revision())
+		})
+	}
+}
+
+// Every override mutation races one seal on a real bucket; afterwards the
+// record is sealed and stays sealed, so no success among them could have
+// landed after the seal.
+func TestSealConcurrentMutations(t *testing.T) {
+	svc, _ := newSealService(t)
+	ctx := context.Background()
+	svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+	require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator"))
+
+	mutations := mutationTable()
+	errs := make([]error, len(mutations))
+	var wg sync.WaitGroup
+	for i, m := range mutations {
+		wg.Add(1)
+		go func(i int, m mutationCase) {
+			defer wg.Done()
+			errs[i] = m.call(svc, ctx)
+		}(i, m)
+	}
+
+	var sealResult *OverrideSeal
+	var sealErr error
+	wg.Go(func() {
+		sealResult, sealErr = svc.SealAccountOverrides(ctx, testAccount)
+	})
+	wg.Wait()
+
+	require.NoError(t, sealErr)
+	require.NotNil(t, sealResult)
+	for i, err := range errs {
+		if err != nil {
+			assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorBlocked), "mutation %s: got %v", mutations[i].name, err)
+		}
+	}
+
+	entry, err := svc.overrides.Get(ctx, testAccount)
+	require.NoError(t, err)
+	assert.Equal(t, OverrideStateSealed, decodeOverride(t, entry).State)
+	assert.Equal(t, sealResult.Revision, entry.Revision())
+
+	_, _, err = svc.GetAccountQuota(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorBlocked))
+}
+
+// A Get pinned to a revision the record has already moved past exhausts the
+// mutation's retries without ever writing.
+func TestMutationStaleReadExhaustsWithoutWriting(t *testing.T) {
+	svc, js := newSealService(t)
+	ctx := context.Background()
+
+	require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator"))
+	stale, err := svc.overrides.Get(ctx, testAccount)
+	require.NoError(t, err)
+
+	require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(16)}, "operator"))
+
+	before := lastOverrideSeq(t, js)
+	svc.overrides = &staleGetKV{KeyValue: svc.overrides, entry: stale}
+
+	err = svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(32)}, "operator")
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInternalError), "got %v", err)
+	assert.Equal(t, before, lastOverrideSeq(t, js))
+}
+
+// A Get that always reports absent, while a sealed record actually exists,
+// never gets to recreate it: every Create attempt conflicts.
+func TestMutationStaleAbsentDoesNotRecreate(t *testing.T) {
+	svc, js := newSealService(t)
+	ctx := context.Background()
+	svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+	_, err := svc.SealAccountOverrides(ctx, testAccount)
+	require.NoError(t, err)
+
+	before := lastOverrideSeq(t, js)
+	svc.overrides = &staleNotFoundKV{KeyValue: svc.overrides}
+
+	err = svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator")
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInternalError), "got %v", err)
+	assert.Equal(t, before, lastOverrideSeq(t, js))
+}
+
+// A revision that keeps moving under sustained concurrent writes must not let
+// the seal spin forever: it gives up, bounded by its attempt budget, and its
+// own write never lands.
+func TestSealSustainedConflictFailsWithoutWriting(t *testing.T) {
+	svc, _ := newSealService(t)
+	ctx := context.Background()
+	svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+	require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator"))
+
+	realKV := svc.overrides
+	race := &sealRaceKV{KeyValue: realKV}
+	race.beforeWrite = func(int) error {
+		data, merr := json.Marshal(&overrideRecord{Overrides: Overrides{VCPUs: ptr(16)}})
+		require.NoError(t, merr)
+		_, perr := realKV.Put(ctx, testAccount, data)
+		return perr
+	}
+	svc.overrides = race
+
+	_, err := svc.SealAccountOverrides(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInternalError), "got %v", err)
+	assert.Equal(t, sealMaxAttempts, race.writeCalls)
+
+	entry, err := realKV.Get(ctx, testAccount)
+	require.NoError(t, err)
+	assert.NotEqual(t, OverrideStateSealed, decodeOverride(t, entry).State, "the seal's own write never landed")
+}
+
+// A store error on the seal's read reports failed, never written.
+func TestSealStoreErrorWritesNothing(t *testing.T) {
+	svc, js := newSealService(t)
+	ctx := context.Background()
+	svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+
+	before := lastOverrideSeq(t, js)
+	svc.overrides = &errGetKV{KeyValue: svc.overrides, err: errors.New("kv unavailable")}
+
+	_, err := svc.SealAccountOverrides(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInternalError), "got %v", err)
+	assert.Equal(t, before, lastOverrideSeq(t, js))
+}
+
+func TestIsRevisionConflictRecognisesTombstoneRetry(t *testing.T) {
+	assert.True(t, isRevisionConflict(jetstream.ErrKeyExists))
+	assert.True(t, isRevisionConflict(jetstream.ErrKeyRevisionMismatch))
+	assert.True(t, isRevisionConflict(&jetstream.APIError{ErrorCode: jetstream.JSErrCodeStreamWrongLastSequence}))
+	assert.True(t, isRevisionConflict(&jetstream.APIError{ErrorCode: jetstream.JSErrCodeStreamWrongLastSequenceConstant}))
+	assert.True(t, isRevisionConflict(fmt.Errorf("wrapped: %w", &jetstream.APIError{ErrorCode: jetstream.JSErrCodeStreamWrongLastSequence})))
+	assert.False(t, isRevisionConflict(errors.New("unrelated failure")))
+	assert.False(t, isRevisionConflict(nil))
+}
+
+// Clearing never writes when there is nothing active to clear: an absent key,
+// a tombstoned key, and a sealed key hidden behind a stale not-found Get all
+// leave the stream exactly where they found it.
+func TestClearOnMissingKeyWritesNothing(t *testing.T) {
+	t.Run("absent", func(t *testing.T) {
+		svc, js := newSealService(t)
+		ctx := context.Background()
+		before := lastOverrideSeq(t, js)
+
+		require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{}, "operator"))
+		assert.Equal(t, before, lastOverrideSeq(t, js))
+	})
+
+	t.Run("tombstone", func(t *testing.T) {
+		svc, js := newSealService(t)
+		ctx := context.Background()
+		require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator"))
+		require.NoError(t, svc.overrides.Delete(ctx, testAccount))
+		before := lastOverrideSeq(t, js)
+
+		require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{}, "operator"))
+		assert.Equal(t, before, lastOverrideSeq(t, js))
+	})
+
+	t.Run("sealed_hidden_by_stale_not_found", func(t *testing.T) {
+		svc, js := newSealService(t)
+		ctx := context.Background()
+		svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+		sealed, err := svc.SealAccountOverrides(ctx, testAccount)
+		require.NoError(t, err)
+		before := lastOverrideSeq(t, js)
+
+		realKV := svc.overrides
+		svc.overrides = &staleNotFoundKV{KeyValue: realKV}
+
+		require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{}, "operator"))
+		assert.Equal(t, before, lastOverrideSeq(t, js))
+
+		entry, err := realKV.Get(ctx, testAccount)
+		require.NoError(t, err)
+		rec := decodeOverride(t, entry)
+		assert.Equal(t, OverrideStateSealed, rec.State)
+		assert.Equal(t, sealed.Revision, entry.Revision())
+	})
+}
+
+// A seal against an absent or a tombstoned key creates the sealed record.
+func TestSealOnMissingKeyCreatesSealedRecord(t *testing.T) {
+	for _, state := range []string{"absent", "tombstone"} {
+		t.Run(state, func(t *testing.T) {
+			svc, _ := newSealService(t)
+			ctx := context.Background()
+			svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+			if state == "tombstone" {
+				require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator"))
+				require.NoError(t, svc.overrides.Delete(ctx, testAccount))
+			}
+			_, err := svc.overrides.Get(ctx, testAccount)
+			require.ErrorIs(t, err, jetstream.ErrKeyNotFound)
+
+			seal, err := svc.SealAccountOverrides(ctx, testAccount)
+			require.NoError(t, err)
+
+			entry, err := svc.overrides.Get(ctx, testAccount)
+			require.NoError(t, err)
+			assert.Equal(t, seal.Revision, entry.Revision())
+			assert.Equal(t, OverrideStateSealed, decodeOverride(t, entry).State)
+		})
+	}
+}
+
+// alwaysConflictThenErrorKV answers Get as absent and every Create attempt
+// with a conflict, then a store error on the final attempt: a recording
+// wrapper proving the seal tried a real write on every attempt before giving
+// up, rather than ever reporting success from an observed absence.
+type alwaysConflictThenErrorKV struct {
+	jetstream.KeyValue
+
+	calls int
+}
+
+func (f *alwaysConflictThenErrorKV) Get(context.Context, string) (jetstream.KeyValueEntry, error) {
+	return nil, jetstream.ErrKeyNotFound
+}
+
+func (f *alwaysConflictThenErrorKV) Create(context.Context, string, []byte, ...jetstream.KVCreateOpt) (uint64, error) {
+	f.calls++
+	if f.calls < sealMaxAttempts {
+		return 0, jetstream.ErrKeyExists
+	}
+	return 0, errors.New("store unavailable")
+}
+
+func TestSealNeverSucceedsWithoutAWrite(t *testing.T) {
+	svc, _ := newSealService(t)
+	ctx := context.Background()
+	svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+	fake := &alwaysConflictThenErrorKV{}
+	svc.overrides = fake
+
+	seal, err := svc.SealAccountOverrides(ctx, testAccount)
+	assert.Error(t, err)
+	assert.Nil(t, seal)
+	assert.Equal(t, sealMaxAttempts, fake.calls, "every attempt must have tried a real write before giving up")
+}
+
+// An account IAM does not hold as TERMINATING refuses the seal before any
+// override-KV write, whether the key is present or absent.
+func TestSealRefusedForActiveAccount(t *testing.T) {
+	for _, state := range []string{"present", "absent"} {
+		t.Run(state, func(t *testing.T) {
+			svc, js := newSealService(t)
+			ctx := context.Background()
+			if state == "present" {
+				require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator"))
+			}
+			svc.SetAccountTerminationChecker(&stubTerminationChecker{terminating: map[string]bool{}})
+
+			before := lastOverrideSeq(t, js)
+			_, err := svc.SealAccountOverrides(ctx, testAccount)
+			assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorIncorrectState), "got %v", err)
+			assert.Equal(t, before, lastOverrideSeq(t, js))
+		})
+	}
+}
+
+func TestSealRefusedForMissingIAMRecord(t *testing.T) {
+	svc, js := newSealService(t)
+	ctx := context.Background()
+	svc.SetAccountTerminationChecker(&stubTerminationChecker{terminating: map[string]bool{}})
+
+	before := lastOverrideSeq(t, js)
+	_, err := svc.SealAccountOverrides(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorIncorrectState), "got %v", err)
+	assert.Equal(t, before, lastOverrideSeq(t, js))
+}
+
+func TestSealRefusedWhenIAMReadFails(t *testing.T) {
+	svc, js := newSealService(t)
+	ctx := context.Background()
+	svc.SetAccountTerminationChecker(&stubTerminationChecker{err: errors.New("iam unavailable")})
+
+	before := lastOverrideSeq(t, js)
+	_, err := svc.SealAccountOverrides(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInternalError), "got %v", err)
+	assert.Equal(t, before, lastOverrideSeq(t, js))
+}
+
+func TestSealRefusedWithNoCheckerWired(t *testing.T) {
+	svc, js := newSealService(t)
+	ctx := context.Background()
+
+	before := lastOverrideSeq(t, js)
+	_, err := svc.SealAccountOverrides(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInternalError), "got %v", err)
+	assert.Equal(t, before, lastOverrideSeq(t, js))
+}
+
+func TestSealAcceptedForTerminatingAccount(t *testing.T) {
+	svc, _ := newSealService(t)
+	ctx := context.Background()
+	svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+
+	seal, err := svc.SealAccountOverrides(ctx, testAccount)
+	require.NoError(t, err)
+	assert.NotZero(t, seal.Revision)
+
+	entry, err := svc.overrides.Get(ctx, testAccount)
+	require.NoError(t, err)
+	assert.Equal(t, seal.Revision, entry.Revision())
+}
+
+// A refused seal, on a gateway that disabled quotas and so holds no override
+// bucket, never creates one.
+func TestSealRefusalDoesNotCreateBucket(t *testing.T) {
+	t.Run("not_terminating", func(t *testing.T) {
+		svc, js := newDisabledSealService(t)
+		ctx := context.Background()
+		svc.SetAccountTerminationChecker(&stubTerminationChecker{terminating: map[string]bool{}})
+
+		_, err := svc.SealAccountOverrides(ctx, testAccount)
+		assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorIncorrectState), "got %v", err)
+		_, err = js.KeyValue(context.Background(), KVBucketAccountQuota)
+		assert.ErrorIs(t, err, jetstream.ErrBucketNotFound)
+	})
+
+	t.Run("iam_read_fails", func(t *testing.T) {
+		svc, js := newDisabledSealService(t)
+		ctx := context.Background()
+		svc.SetAccountTerminationChecker(&stubTerminationChecker{err: errors.New("iam unavailable")})
+
+		_, err := svc.SealAccountOverrides(ctx, testAccount)
+		assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInternalError), "got %v", err)
+		_, err = js.KeyValue(context.Background(), KVBucketAccountQuota)
+		assert.ErrorIs(t, err, jetstream.ErrBucketNotFound)
+	})
+}
+
+// A seal accepted on a gateway whose quotas were disabled at startup creates
+// the override bucket at seal time, never at startup.
+func TestSealOnQuotaDisabledGatewayCreatesBucket(t *testing.T) {
+	svc, js := newDisabledSealService(t)
+	ctx := context.Background()
+	svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+
+	seal, err := svc.SealAccountOverrides(ctx, testAccount)
+	require.NoError(t, err)
+
+	kv, err := js.KeyValue(context.Background(), KVBucketAccountQuota)
+	require.NoError(t, err)
+	entry, err := kv.Get(context.Background(), testAccount)
+	require.NoError(t, err)
+	assert.Equal(t, seal.Revision, entry.Revision())
+	assert.Equal(t, OverrideStateSealed, decodeOverride(t, entry).State)
+}
+
+// A sealed record stays sealed after the service is rebuilt over the same
+// JetStream, every read and write refuses it, and a re-seal gets a higher
+// revision.
+func TestSealSurvivesServiceReconstruction(t *testing.T) {
+	_, _, js := testutil.StartTestJetStream(t)
+	ctx := context.Background()
+
+	kv, err := js.CreateKeyValue(ctx, jetstream.KeyValueConfig{Bucket: KVBucketAccountQuota, History: 1})
+	require.NoError(t, err)
+
+	svc := New(baseLimits(), nil)
+	svc.SetOverrides(kv)
+	svc.SetJetStream(js)
+	svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+	first, err := svc.SealAccountOverrides(ctx, testAccount)
+	require.NoError(t, err)
+
+	rebuiltKV, err := js.KeyValue(ctx, KVBucketAccountQuota)
+	require.NoError(t, err)
+	rebuilt := New(baseLimits(), nil)
+	rebuilt.SetOverrides(rebuiltKV)
+	rebuilt.SetJetStream(js)
+
+	entry, err := rebuilt.overrides.Get(ctx, testAccount)
+	require.NoError(t, err)
+	assert.Equal(t, OverrideStateSealed, decodeOverride(t, entry).State)
+
+	_, _, err = rebuilt.GetAccountQuota(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorBlocked))
+	assert.True(t, awserrors.IsErrorCode(rebuilt.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator"), awserrors.ErrorBlocked))
+	assert.True(t, awserrors.IsErrorCode(rebuilt.PutAccountQuota(ctx, testAccount, Overrides{}, "operator"), awserrors.ErrorBlocked))
+
+	limits, err := rebuilt.limitsFor(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorBlocked))
+	assert.Zero(t, limits)
+
+	rebuilt.SetAccountTerminationChecker(terminatingChecker(testAccount))
+	second, err := rebuilt.SealAccountOverrides(ctx, testAccount)
+	require.NoError(t, err)
+	assert.Greater(t, second.Revision, first.Revision)
+}
+
+func TestSealRefusesEmptyInvalidAndGlobalAccount(t *testing.T) {
+	svc, _ := newSealService(t)
+	ctx := context.Background()
+
+	_, err := svc.SealAccountOverrides(ctx, "")
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorMissingParameter), "got %v", err)
+
+	_, err = svc.SealAccountOverrides(ctx, "not-an-account")
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInvalidParameterValue), "got %v", err)
+
+	_, err = svc.SealAccountOverrides(ctx, awsidentifiers.GlobalAccountID)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInvalidParameterValue), "got %v", err)
+}
+
+func TestSealLeavesOtherAccountsUntouched(t *testing.T) {
+	svc, _ := newSealService(t)
+	ctx := context.Background()
+	accountB := "222222222222"
+	require.NoError(t, svc.PutAccountQuota(ctx, accountB, Overrides{VCPUs: ptr(8)}, "operator"))
+
+	svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+	_, err := svc.SealAccountOverrides(ctx, testAccount)
+	require.NoError(t, err)
+
+	over, _, err := svc.GetAccountQuota(ctx, accountB)
+	require.NoError(t, err)
+	require.NotNil(t, over.VCPUs)
+	assert.Equal(t, 8, *over.VCPUs)
+}
+
+// A record written before the state field existed has none and is treated as
+// active by both reads and writes; a write on it preserves the old shape.
+func TestRecordWithoutStateIsActive(t *testing.T) {
+	svc, _ := newSealService(t)
+	ctx := context.Background()
+
+	data, err := json.Marshal(Overrides{VCPUs: ptr(8)})
+	require.NoError(t, err)
+	_, err = svc.overrides.Put(ctx, testAccount, data)
+	require.NoError(t, err)
+
+	over, limits, err := svc.GetAccountQuota(ctx, testAccount)
+	require.NoError(t, err)
+	require.NotNil(t, over.VCPUs)
+	assert.Equal(t, 8, limits.VCPUs)
+
+	require.NoError(t, svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(16)}, "operator"))
+	entry, err := svc.overrides.Get(ctx, testAccount)
+	require.NoError(t, err)
+	assert.NotContains(t, string(entry.Value()), `"state"`)
+}
+
+// A record whose state is neither active nor sealed fails every read and
+// write closed, rather than being treated as active.
+func TestUnknownStateFailsClosed(t *testing.T) {
+	svc, _ := newSealService(t)
+	ctx := context.Background()
+
+	_, err := svc.overrides.Put(ctx, testAccount, []byte(`{"state":"frozen"}`))
+	require.NoError(t, err)
+
+	_, _, err = svc.GetAccountQuota(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInternalError), "got %v", err)
+
+	limits, err := svc.limitsFor(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorInternalError), "got %v", err)
+	assert.Zero(t, limits)
+
+	assert.True(t, awserrors.IsErrorCode(svc.PutAccountQuota(ctx, testAccount, Overrides{VCPUs: ptr(8)}, "operator"), awserrors.ErrorInternalError))
+	assert.True(t, awserrors.IsErrorCode(svc.PutAccountQuota(ctx, testAccount, Overrides{}, "operator"), awserrors.ErrorInternalError))
+}
+
+// A sealed record is Blocked for admission's own resolver, which every
+// caller (admin read, ELBv2 DescribeAccountLimits, every Enforce* method)
+// shares.
+func TestAdmissionOnSealedRecordIsBlocked(t *testing.T) {
+	svc, _ := newSealService(t)
+	ctx := context.Background()
+	svc.SetAccountTerminationChecker(terminatingChecker(testAccount))
+	_, err := svc.SealAccountOverrides(ctx, testAccount)
+	require.NoError(t, err)
+
+	limits, err := svc.limitsFor(ctx, testAccount)
+	assert.True(t, awserrors.IsErrorCode(err, awserrors.ErrorBlocked), "got %v", err)
+	assert.Zero(t, limits)
+
+	assert.Error(t, svc.CheckVCPU(ctx, testAccount, 1))
+	assert.True(t, awserrors.IsErrorCode(svc.CheckVCPU(ctx, testAccount, 1), awserrors.ErrorBlocked))
+}
+
+// The existing read-failure fallback in limitsFor (a decode or store error
+// falls back to the configured limits) is unchanged by the seal work: it is
+// distinct from a successfully read sealed record, which refuses instead.
+func TestAdmissionReadFailureFallbackUnchanged(t *testing.T) {
+	svc, kv := newOverrideService(t)
+
+	_, err := kv.Put(t.Context(), testAccount, []byte("{not json"))
+	require.NoError(t, err)
+
+	limits, err := svc.limitsFor(t.Context(), testAccount)
+	require.NoError(t, err, "a decode failure must still fall back, not error, from limitsFor")
+	require.Equal(t, baseLimits(), limits)
+}

spinifex/domains/admission/quota/storage.go

@@ -23,7 +23,10 @@ func (s *Service) EnforceVolumeCreate(ctx context.Context, natsConn *nats.Conn,
 	if err != nil {
 		return err
 	}
-	limits := s.limitsFor(ctx, accountID)
+	limits, err := s.limitsFor(ctx, accountID)
+	if err != nil {
+		return err
+	}
 	if err := exceeds(len(out.Volumes), 1, limits.Volumes); err != nil {
 		return err
 	}
@@ -43,7 +46,11 @@ func (s *Service) EnforceVolumeModify(ctx context.Context, natsConn *nats.Conn,
 	if err != nil {
 		return err
 	}
-	return exceeds(total-oldGiB, newGiB, s.limitsFor(ctx, accountID).VolumesGiB)
+	limits, err := s.limitsFor(ctx, accountID)
+	if err != nil {
+		return err
+	}
+	return exceeds(total-oldGiB, newGiB, limits.VolumesGiB)
 }
 
 // volumeUsage sums the account's volume sizes in GiB via a single DescribeVolumes

spinifex/domains/admission/quota/vcpu.go

@@ -37,7 +37,11 @@ func (s *Service) CheckVCPU(ctx context.Context, accountID string, want int) err
 	if err != nil {
 		return err
 	}
-	return exceeds(current, want, s.limitsFor(ctx, accountID).VCPUs)
+	limits, err := s.limitsFor(ctx, accountID)
+	if err != nil {
+		return err
+	}
+	return exceeds(current, want, limits.VCPUs)
 }
 
 // casVCPU applies next to accountID's counter under bounded JetStream CAS so

spinifex/domains/ec2/awsapi/instance/DescribeInstanceTypes.go

@@ -3,7 +3,6 @@ package instance
 import (
 	"context"
 	"encoding/json"
-	"errors"
 	"fmt"
 	"github.com/mulgadc/spinifex/spinifex/foundation/messaging/nats"
 	"log/slog"
@@ -101,10 +100,10 @@ func DescribeInstanceTypes(ctx context.Context, input *ec2.DescribeInstanceTypes
 
 		if missing := missingInstanceTypes(finalInstanceTypes, input.InstanceTypes); len(missing) > 0 {
 			if !proveAbsence || !everyNodeAnswered(nodeIDs, answered, sum) {
-				return nil, errors.Join(
+				return nil, awserrors.WithRetryAfter(
 					awserrors.Errorf(awserrors.ErrorServiceUnavailable,
 						"cannot confirm instance type(s) %s: describe fan-out did not complete", strings.Join(missing, ", ")),
-					awserrors.RetryAfter(awserrors.ErrorServiceUnavailable, describeIncompleteRetryAfter),
+					describeIncompleteRetryAfter,
 				)
 			}
 			return nil, awserrors.Errorf(awserrors.ErrorInvalidInstanceType,

spinifex/domains/ec2/awsapi/instance/DescribeInstances.go

@@ -14,8 +14,8 @@ import (
 
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/ec2"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 )
 
@@ -126,10 +126,10 @@ func DescribeInstancesChecked(ctx context.Context, input *ec2.DescribeInstancesI
 			// An incomplete sweep cannot prove absence, so a missing id is
 			// unresolved, not gone. NotFound would be a false negative and an
 			// empty success would look identical to a real one, so return 503.
-			return nil, errors.Join(
+			return nil, awserrors.WithRetryAfter(
 				awserrors.Errorf(awserrors.ErrorServiceUnavailable,
 					"cannot confirm status of instance(s) %s: describe fan-out did not complete", strings.Join(missing, ", ")),
-				awserrors.RetryAfter(awserrors.ErrorServiceUnavailable, describeIncompleteRetryAfter),
+				describeIncompleteRetryAfter,
 			)
 		}
 	}
@@ -198,10 +198,10 @@ func DescribeAccountInstancesComplete(ctx context.Context, natsConn *nats.Conn,
 }
 
 func incompleteListingError() error {
-	return errors.Join(
+	return awserrors.WithRetryAfter(
 		awserrors.Errorf(awserrors.ErrorServiceUnavailable,
 			"cannot list instances: not every node and instance store answered"),
-		awserrors.RetryAfter(awserrors.ErrorServiceUnavailable, describeIncompleteRetryAfter),
+		describeIncompleteRetryAfter,
 	)
 }
 
@@ -477,7 +477,7 @@ func judgeIdentityCompleteness(ctx context.Context, frames []natsmsg.Frame, sum
 // that carries only an ARN. Results are cached per ARN to avoid repeated RPCs.
 // Misses are warn-logged and leave Id empty (graceful degradation for deleted profiles).
 // Safe to call with a nil output or nil IAMService (no-op).
-func EnrichInstanceProfileIDs(out *ec2.DescribeInstancesOutput, iamSvc handlers_iam.IAMService, accountID string) {
+func EnrichInstanceProfileIDs(out *ec2.DescribeInstancesOutput, iamSvc iamdomain.IAMService, accountID string) {
 	if out == nil || iamSvc == nil {
 		return
 	}

spinifex/domains/ec2/awsapi/instance/DescribeInstances_test.go

@@ -11,8 +11,8 @@ import (
 
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/ec2"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -751,9 +751,9 @@ func TestEnrichInstanceProfileIDs_CachesPerARN(t *testing.T) {
 	arn := "arn:aws:iam::123456789012:instance-profile/shared"
 	var calls int
 	svc := &fakeIAMService{
-		resolveFn: func(_, nameOrARN string) (*handlers_iam.InstanceProfile, error) {
+		resolveFn: func(_, nameOrARN string) (*iamdomain.InstanceProfile, error) {
 			calls++
-			return &handlers_iam.InstanceProfile{ARN: nameOrARN, InstanceProfileID: "AIPAEXAMPLE"}, nil
+			return &iamdomain.InstanceProfile{ARN: nameOrARN, InstanceProfileID: "AIPAEXAMPLE"}, nil
 		},
 	}
 
@@ -771,11 +771,11 @@ func TestEnrichInstanceProfileIDs_ResolveErrorLeavesIdEmpty(t *testing.T) {
 	failingARN := "arn:aws:iam::123456789012:instance-profile/missing"
 	okARN := "arn:aws:iam::123456789012:instance-profile/present"
 	svc := &fakeIAMService{
-		resolveFn: func(_, nameOrARN string) (*handlers_iam.InstanceProfile, error) {
+		resolveFn: func(_, nameOrARN string) (*iamdomain.InstanceProfile, error) {
 			if nameOrARN == failingARN {
 				return nil, errors.New(awserrors.ErrorIAMNoSuchEntity)
 			}
-			return &handlers_iam.InstanceProfile{ARN: nameOrARN, InstanceProfileID: "AIPAOK"}, nil
+			return &iamdomain.InstanceProfile{ARN: nameOrARN, InstanceProfileID: "AIPAOK"}, nil
 		},
 	}
 

spinifex/domains/ec2/awsapi/instance/IamInstanceProfileAssociation.go

@@ -12,9 +12,9 @@ import (
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/ec2"
 	"github.com/mulgadc/spinifex/contracts/ec2/v1"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/arn"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 )
 
@@ -24,7 +24,7 @@ const fanOutTimeout = 3 * time.Second
 
 // resolveAndAuthorizeProfile resolves a profile (name or ARN) and enforces iam:PassRole.
 // Profiles without a role skip PassRole; missing profile → InvalidIamInstanceProfile.NotFound.
-func resolveAndAuthorizeProfile(spec *ec2.IamInstanceProfileSpecification, iamSvc handlers_iam.IAMService, accountID string, passRoleCheck PassRoleChecker) (*handlers_iam.InstanceProfile, error) {
+func resolveAndAuthorizeProfile(spec *ec2.IamInstanceProfileSpecification, iamSvc iamdomain.IAMService, accountID string, passRoleCheck PassRoleChecker) (*iamdomain.InstanceProfile, error) {
 	nameOrARN := profileNameOrARN(spec)
 	if nameOrARN == "" {
 		return nil, errors.New(awserrors.ErrorMissingParameter)
@@ -55,7 +55,7 @@ func resolveAndAuthorizeProfile(spec *ec2.IamInstanceProfileSpecification, iamSv
 // AssociateIamInstanceProfile attaches a profile to a running instance.
 // The gateway resolves the profile and enforces PassRole; the daemon validates
 // no existing profile and generates a fresh iip-assoc-… ID.
-func AssociateIamInstanceProfile(ctx context.Context, input *ec2.AssociateIamInstanceProfileInput, natsConn *nats.Conn, iamSvc handlers_iam.IAMService, accountID string, passRoleCheck PassRoleChecker) (*ec2.AssociateIamInstanceProfileOutput, error) {
+func AssociateIamInstanceProfile(ctx context.Context, input *ec2.AssociateIamInstanceProfileInput, natsConn *nats.Conn, iamSvc iamdomain.IAMService, accountID string, passRoleCheck PassRoleChecker) (*ec2.AssociateIamInstanceProfileOutput, error) {
 	if input == nil || input.InstanceId == nil || *input.InstanceId == "" {
 		return nil, errors.New(awserrors.ErrorMissingParameter)
 	}
@@ -107,7 +107,7 @@ func DisassociateIamInstanceProfile(ctx context.Context, input *ec2.Disassociate
 
 // ReplaceIamInstanceProfileAssociation swaps the profile for an existing association.
 // The gateway resolves the new profile and enforces PassRole; the daemon swaps atomically.
-func ReplaceIamInstanceProfileAssociation(ctx context.Context, input *ec2.ReplaceIamInstanceProfileAssociationInput, natsConn *nats.Conn, iamSvc handlers_iam.IAMService, expectedNodes int, accountID string, passRoleCheck PassRoleChecker) (*ec2.ReplaceIamInstanceProfileAssociationOutput, error) {
+func ReplaceIamInstanceProfileAssociation(ctx context.Context, input *ec2.ReplaceIamInstanceProfileAssociationInput, natsConn *nats.Conn, iamSvc iamdomain.IAMService, expectedNodes int, accountID string, passRoleCheck PassRoleChecker) (*ec2.ReplaceIamInstanceProfileAssociationOutput, error) {
 	if input == nil || input.AssociationId == nil || *input.AssociationId == "" {
 		return nil, errors.New(awserrors.ErrorMissingParameter)
 	}
@@ -173,7 +173,7 @@ func DescribeIamInstanceProfileAssociations(ctx context.Context, input *ec2.Desc
 
 // enrichProfileID fills IamInstanceProfile.Id from the resolved profile.
 // Daemons have no IAM access. Safe to call with a nil association.
-func enrichProfileID(assoc *ec2.IamInstanceProfileAssociation, profile *handlers_iam.InstanceProfile) {
+func enrichProfileID(assoc *ec2.IamInstanceProfileAssociation, profile *iamdomain.InstanceProfile) {
 	if assoc == nil || profile == nil {
 		return
 	}

spinifex/domains/ec2/awsapi/instance/IamInstanceProfileAssociation_test.go

@@ -11,10 +11,10 @@ import (
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/ec2"
 	"github.com/mulgadc/spinifex/contracts/ec2/v1"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/arn"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
 	awsidentifiers "github.com/mulgadc/spinifex/spinifex/foundation/aws/identifiers"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -35,13 +35,13 @@ const (
 // fakeIAMService implements profile and canonical role lookups.
 // Other methods nil-panic to surface unexpected calls.
 type fakeIAMService struct {
-	handlers_iam.IAMService
+	iamdomain.IAMService
 
-	resolveFn   func(accountID, nameOrARN string) (*handlers_iam.InstanceProfile, error)
+	resolveFn   func(accountID, nameOrARN string) (*iamdomain.InstanceProfile, error)
 	canonicalFn func(accountID string, kind arn.IAMResourceType, name string) (string, error)
 }
 
-func (f *fakeIAMService) ResolveInstanceProfile(accountID, nameOrARN string) (*handlers_iam.InstanceProfile, error) {
+func (f *fakeIAMService) ResolveInstanceProfile(accountID, nameOrARN string) (*iamdomain.InstanceProfile, error) {
 	if f.resolveFn != nil {
 		return f.resolveFn(accountID, nameOrARN)
 	}
@@ -58,11 +58,11 @@ func (f *fakeIAMService) CanonicalResourceARN(accountID string, kind arn.IAMReso
 	return "", errors.New(awserrors.ErrorIAMNoSuchEntity)
 }
 
-var _ handlers_iam.IAMService = (*fakeIAMService)(nil)
+var _ iamdomain.IAMService = (*fakeIAMService)(nil)
 
 // profileWithRole builds a resolved profile fixture pointing at the test role.
-func profileWithRole() *handlers_iam.InstanceProfile {
-	return &handlers_iam.InstanceProfile{
+func profileWithRole() *iamdomain.InstanceProfile {
+	return &iamdomain.InstanceProfile{
 		InstanceProfileName: testProfileNameApp,
 		InstanceProfileID:   testProfileIDApp,
 		AccountID:           testGwAccountID,
@@ -73,8 +73,8 @@ func profileWithRole() *handlers_iam.InstanceProfile {
 
 // profileNoRole builds a resolved profile fixture with no role attached.
 // resolveAndAuthorizeProfile must NOT invoke PassRole for this profile.
-func profileNoRole() *handlers_iam.InstanceProfile {
-	return &handlers_iam.InstanceProfile{
+func profileNoRole() *iamdomain.InstanceProfile {
+	return &iamdomain.InstanceProfile{
 		InstanceProfileName: testProfileNameOther,
 		InstanceProfileID:   testProfileIDOther,
 		AccountID:           testGwAccountID,
@@ -88,7 +88,7 @@ func profileNoRole() *handlers_iam.InstanceProfile {
 func TestResolveAndAuthorizeProfile_NameForm(t *testing.T) {
 	t.Parallel()
 	gotNameOrARN := ""
-	svc := &fakeIAMService{resolveFn: func(_, nameOrARN string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(_, nameOrARN string) (*iamdomain.InstanceProfile, error) {
 		gotNameOrARN = nameOrARN
 		return profileWithRole(), nil
 	}}
@@ -112,7 +112,7 @@ func TestResolveAndAuthorizeProfile_NameForm(t *testing.T) {
 func TestResolveAndAuthorizeProfile_ArnForm(t *testing.T) {
 	t.Parallel()
 	gotNameOrARN := ""
-	svc := &fakeIAMService{resolveFn: func(_, nameOrARN string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(_, nameOrARN string) (*iamdomain.InstanceProfile, error) {
 		gotNameOrARN = nameOrARN
 		return profileWithRole(), nil
 	}}
@@ -144,7 +144,7 @@ func TestResolveAndAuthorizeProfile_NilIAMService(t *testing.T) {
 
 func TestResolveAndAuthorizeProfile_NotFoundMapsToEC2Error(t *testing.T) {
 	t.Parallel()
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return nil, errors.New(awserrors.ErrorIAMNoSuchEntity)
 	}}
 	_, err := resolveAndAuthorizeProfile(
@@ -157,7 +157,7 @@ func TestResolveAndAuthorizeProfile_NotFoundMapsToEC2Error(t *testing.T) {
 
 func TestResolveAndAuthorizeProfile_PassthroughOtherIAMError(t *testing.T) {
 	t.Parallel()
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return nil, errors.New(awserrors.ErrorAccessDenied)
 	}}
 	_, err := resolveAndAuthorizeProfile(
@@ -170,7 +170,7 @@ func TestResolveAndAuthorizeProfile_PassthroughOtherIAMError(t *testing.T) {
 
 func TestResolveAndAuthorizeProfile_PassRoleDenied(t *testing.T) {
 	t.Parallel()
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileWithRole(), nil
 	}}
 	check := func(string) error { return errors.New(awserrors.ErrorAccessDenied) }
@@ -183,7 +183,7 @@ func TestResolveAndAuthorizeProfile_PassRoleDenied(t *testing.T) {
 
 func TestResolveAndAuthorizeProfile_NoRoleSkipsPassRole(t *testing.T) {
 	t.Parallel()
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileNoRole(), nil
 	}}
 	checkCalled := false
@@ -265,7 +265,7 @@ func TestAssociateIamInstanceProfile_ProfileNotFound(t *testing.T) {
 
 func TestAssociateIamInstanceProfile_PassRoleDenied(t *testing.T) {
 	t.Parallel()
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileWithRole(), nil
 	}}
 	in := &ec2.AssociateIamInstanceProfileInput{
@@ -281,7 +281,7 @@ func TestAssociateIamInstanceProfile_PassRoleDenied(t *testing.T) {
 func TestAssociateIamInstanceProfile_NoResponders(t *testing.T) {
 	t.Parallel()
 	_, nc := startTestNATSServer(t)
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileNoRole(), nil
 	}}
 	in := &ec2.AssociateIamInstanceProfileInput{
@@ -299,7 +299,7 @@ func TestAssociateIamInstanceProfile_NoResponders(t *testing.T) {
 func TestAssociateIamInstanceProfile_Success(t *testing.T) {
 	t.Parallel()
 	_, nc := startTestNATSServer(t)
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileWithRole(), nil
 	}}
 	const instanceID = "i-assoc-success"
@@ -345,7 +345,7 @@ func TestAssociateIamInstanceProfile_Success(t *testing.T) {
 func TestAssociateIamInstanceProfile_DaemonAlreadyAssociated(t *testing.T) {
 	t.Parallel()
 	_, nc := startTestNATSServer(t)
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileNoRole(), nil
 	}}
 	const instanceID = "i-already-bound"
@@ -474,7 +474,7 @@ func TestReplaceIamInstanceProfileAssociation_MissingProfileSpec(t *testing.T) {
 func TestReplaceIamInstanceProfileAssociation_PassRoleDeniedBeforeBroadcast(t *testing.T) {
 	t.Parallel()
 	// nc=nil proves the call is rejected before any NATS dispatch.
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileWithRole(), nil
 	}}
 	check := func(string) error { return errors.New(awserrors.ErrorAccessDenied) }
@@ -489,7 +489,7 @@ func TestReplaceIamInstanceProfileAssociation_PassRoleDeniedBeforeBroadcast(t *t
 func TestReplaceIamInstanceProfileAssociation_Success(t *testing.T) {
 	t.Parallel()
 	_, nc := startTestNATSServer(t)
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileWithRole(), nil
 	}}
 	const oldID = "iip-assoc-old00000000001"
@@ -530,7 +530,7 @@ func TestReplaceIamInstanceProfileAssociation_Success(t *testing.T) {
 func TestReplaceIamInstanceProfileAssociation_NoSuchAssociation(t *testing.T) {
 	t.Parallel()
 	_, nc := startTestNATSServer(t)
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileNoRole(), nil
 	}}
 	_, err := nc.Subscribe("ec2.IamProfileAssociation.replace", func(msg *nats.Msg) {
@@ -700,7 +700,7 @@ func TestResolveAndAuthorizeProfile_CanonicalRoleARN(t *testing.T) {
 			t.Parallel()
 			storedARN := "arn:aws:iam::111122223333:role" + path + "app-role"
 			svc := &fakeIAMService{
-				resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+				resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 					return profileWithRole(), nil
 				},
 				canonicalFn: func(accountID string, kind arn.IAMResourceType, name string) (string, error) {
@@ -729,7 +729,7 @@ func TestResolveAndAuthorizeProfile_CanonicalLookupError(t *testing.T) {
 		t.Run(lookupErr.Error(), func(t *testing.T) {
 			t.Parallel()
 			svc := &fakeIAMService{
-				resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+				resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 					return profileWithRole(), nil
 				},
 				canonicalFn: func(string, arn.IAMResourceType, string) (string, error) {

spinifex/domains/ec2/awsapi/instance/RunInstances.go

@@ -20,8 +20,8 @@ import (
 	ec2key "github.com/mulgadc/spinifex/spinifex/domains/ec2/key"
 	ec2launchtemplate "github.com/mulgadc/spinifex/spinifex/domains/ec2/launchtemplate"
 	ec2placementgroup "github.com/mulgadc/spinifex/spinifex/domains/ec2/placementgroup"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 )
 
@@ -131,7 +131,7 @@ func hasPinnedPrivateIP(input *ec2.RunInstancesInput) bool {
 // passRoleCheck may be nil to skip PassRole enforcement.
 // launchQuotaCheck may be nil to skip quota enforcement; when present it runs
 // after validation and PassRole authorization, but before any launch dispatch.
-func RunInstances(ctx context.Context, input *ec2.RunInstancesInput, natsConn *nats.Conn, iamSvc handlers_iam.IAMService, accountID string, passRoleCheck PassRoleChecker, launchQuotaCheck LaunchQuotaChecker, expectedNodes int) (reservation ec2.Reservation, err error) {
+func RunInstances(ctx context.Context, input *ec2.RunInstancesInput, natsConn *nats.Conn, iamSvc iamdomain.IAMService, accountID string, passRoleCheck PassRoleChecker, launchQuotaCheck LaunchQuotaChecker, expectedNodes int) (reservation ec2.Reservation, err error) {
 	// Expand a referenced launch template before validation so validation, quota,
 	// and downstream routing all see the effective (post-merge) parameters.
 	if err = ExpandLaunchTemplate(ctx, natsConn, input, accountID); err != nil {
@@ -163,7 +163,7 @@ func RunInstances(ctx context.Context, input *ec2.RunInstancesInput, natsConn *n
 
 // runInstancesInner performs the actual launch: profile resolution, placement
 // routing, and capacity-aware distribution. Wrapped by RunInstances for idempotency.
-func runInstancesInner(ctx context.Context, input *ec2.RunInstancesInput, natsConn *nats.Conn, iamSvc handlers_iam.IAMService, accountID string, passRoleCheck PassRoleChecker, launchQuotaCheck LaunchQuotaChecker, expectedNodes int) (reservation ec2.Reservation, err error) {
+func runInstancesInner(ctx context.Context, input *ec2.RunInstancesInput, natsConn *nats.Conn, iamSvc iamdomain.IAMService, accountID string, passRoleCheck PassRoleChecker, launchQuotaCheck LaunchQuotaChecker, expectedNodes int) (reservation ec2.Reservation, err error) {
 	resolvedProfile, err := resolveAndAuthorizeInstanceProfile(input, iamSvc, accountID, passRoleCheck)
 	if err != nil {
 		return reservation, err
@@ -236,7 +236,7 @@ func runInstancesInner(ctx context.Context, input *ec2.RunInstancesInput, natsCo
 
 // resolveAndAuthorizeInstanceProfile resolves an optional instance profile,
 // enforces PassRole, and normalises input to the canonical ARN only.
-func resolveAndAuthorizeInstanceProfile(input *ec2.RunInstancesInput, iamSvc handlers_iam.IAMService, accountID string, passRoleCheck PassRoleChecker) (*handlers_iam.InstanceProfile, error) {
+func resolveAndAuthorizeInstanceProfile(input *ec2.RunInstancesInput, iamSvc iamdomain.IAMService, accountID string, passRoleCheck PassRoleChecker) (*iamdomain.InstanceProfile, error) {
 	if input.IamInstanceProfile == nil {
 		return nil, nil
 	}
@@ -261,7 +261,7 @@ func profileNameOrARN(spec *ec2.IamInstanceProfileSpecification) string {
 
 // enrichReservationWithProfileID fills IamInstanceProfile.Id on every instance.
 // Daemons emit Arn only (no IAM access); the gateway adds Id from the resolved profile.
-func enrichReservationWithProfileID(reservation *ec2.Reservation, profile *handlers_iam.InstanceProfile) {
+func enrichReservationWithProfileID(reservation *ec2.Reservation, profile *iamdomain.InstanceProfile) {
 	if reservation == nil || profile == nil {
 		return
 	}

spinifex/domains/ec2/awsapi/instance/RunInstances_test.go

@@ -9,8 +9,8 @@ import (
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/ec2"
 	"github.com/mulgadc/spinifex/spinifex/domains/ec2/ebs/metadata"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
@@ -385,7 +385,7 @@ func TestResolveAndAuthorizeInstanceProfile_Absent(t *testing.T) {
 
 func TestResolveAndAuthorizeInstanceProfile_NameForm_NormalisesToARN(t *testing.T) {
 	t.Parallel()
-	svc := &fakeIAMService{resolveFn: func(_, nameOrARN string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(_, nameOrARN string) (*iamdomain.InstanceProfile, error) {
 		assert.Equal(t, testProfileNameApp, nameOrARN)
 		return profileWithRole(), nil
 	}}
@@ -404,7 +404,7 @@ func TestResolveAndAuthorizeInstanceProfile_NameForm_NormalisesToARN(t *testing.
 
 func TestResolveAndAuthorizeInstanceProfile_ArnForm(t *testing.T) {
 	t.Parallel()
-	svc := &fakeIAMService{resolveFn: func(_, nameOrARN string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(_, nameOrARN string) (*iamdomain.InstanceProfile, error) {
 		assert.Equal(t, testProfileARNApp, nameOrARN)
 		return profileWithRole(), nil
 	}}
@@ -433,7 +433,7 @@ func TestResolveAndAuthorizeInstanceProfile_NilIAMService(t *testing.T) {
 
 func TestResolveAndAuthorizeInstanceProfile_NotFoundMapsToEC2Error(t *testing.T) {
 	t.Parallel()
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return nil, errors.New(awserrors.ErrorIAMNoSuchEntity)
 	}}
 	in := runValidInputWithProfile(&ec2.IamInstanceProfileSpecification{Name: aws.String("ghost")})
@@ -445,7 +445,7 @@ func TestResolveAndAuthorizeInstanceProfile_NotFoundMapsToEC2Error(t *testing.T)
 func TestResolveAndAuthorizeInstanceProfile_CrossAccountARNPassthrough(t *testing.T) {
 	t.Parallel()
 	// IAM service rejects cross-account; gateway surfaces the AccessDenied verbatim.
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return nil, errors.New(awserrors.ErrorAccessDenied)
 	}}
 	in := runValidInputWithProfile(&ec2.IamInstanceProfileSpecification{Arn: aws.String(testCrossAccountARN)})
@@ -456,7 +456,7 @@ func TestResolveAndAuthorizeInstanceProfile_CrossAccountARNPassthrough(t *testin
 
 func TestResolveAndAuthorizeInstanceProfile_PassRoleDenied(t *testing.T) {
 	t.Parallel()
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileWithRole(), nil
 	}}
 	check := func(roleARN string) error {
@@ -471,7 +471,7 @@ func TestResolveAndAuthorizeInstanceProfile_PassRoleDenied(t *testing.T) {
 
 func TestResolveAndAuthorizeInstanceProfile_NoRoleSkipsPassRole(t *testing.T) {
 	t.Parallel()
-	svc := &fakeIAMService{resolveFn: func(string, string) (*handlers_iam.InstanceProfile, error) {
+	svc := &fakeIAMService{resolveFn: func(string, string) (*iamdomain.InstanceProfile, error) {
 		return profileNoRole(), nil
 	}}
 	checkCalled := false

spinifex/domains/ec2/awsapi/spotinstance/RequestSpotInstances.go

@@ -17,10 +17,10 @@ import (
 	admissionquota "github.com/mulgadc/spinifex/spinifex/domains/admission/quota"
 	ec2instanceapi "github.com/mulgadc/spinifex/spinifex/domains/ec2/awsapi/instance"
 	ec2spotinstance "github.com/mulgadc/spinifex/spinifex/domains/ec2/spotinstance"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
 	awsidentifiers "github.com/mulgadc/spinifex/spinifex/foundation/aws/identifiers"
 	awstags "github.com/mulgadc/spinifex/spinifex/foundation/aws/tags"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 )
 
@@ -64,7 +64,7 @@ func ValidateRequestSpotInstancesInput(input *ec2.RequestSpotInstancesInput) err
 // ClientToken passed through, same vCPU quota gate), then builds and persists one
 // active/fulfilled SpotInstanceRequest per launched instance. On a launch failure
 // (including InsufficientInstanceCapacity) it returns the error and persists nothing.
-func RequestSpotInstances(ctx context.Context, input *ec2.RequestSpotInstancesInput, natsConn *nats.Conn, iamSvc handlers_iam.IAMService, accountID, az string, passRoleCheck ec2instanceapi.PassRoleChecker, quota *admissionquota.Service, expectedNodes int) (ec2.RequestSpotInstancesOutput, error) {
+func RequestSpotInstances(ctx context.Context, input *ec2.RequestSpotInstancesInput, natsConn *nats.Conn, iamSvc iamdomain.IAMService, accountID, az string, passRoleCheck ec2instanceapi.PassRoleChecker, quota *admissionquota.Service, expectedNodes int) (ec2.RequestSpotInstancesOutput, error) {
 	var output ec2.RequestSpotInstancesOutput
 
 	if err := ValidateRequestSpotInstancesInput(input); err != nil {

spinifex/domains/ec2/guestmetadata/metadata_test.go

@@ -13,8 +13,8 @@ import (
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/iam"
 	"github.com/aws/aws-sdk-go/service/sts"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
@@ -62,13 +62,13 @@ func (f *fakeResolver) resolveVPCCIDR(_ context.Context, _, _ string) (string, e
 }
 
 type fakeIAM struct {
-	profile    *handlers_iam.InstanceProfile
+	profile    *iamdomain.InstanceProfile
 	profileErr error
 	roleARN    string
 	roleErr    error
 }
 
-func (f *fakeIAM) ResolveInstanceProfile(_ context.Context, _, _ string) (*handlers_iam.InstanceProfile, error) {
+func (f *fakeIAM) ResolveInstanceProfile(_ context.Context, _, _ string) (*iamdomain.InstanceProfile, error) {
 	return f.profile, f.profileErr
 }
 
@@ -991,8 +991,8 @@ func TestHTTP_PublicKeysInstanceBackendError500(t *testing.T) {
 
 // ----- IAM / credentials -------------------------------------------------
 
-func profileFixture() *handlers_iam.InstanceProfile {
-	return &handlers_iam.InstanceProfile{
+func profileFixture() *iamdomain.InstanceProfile {
+	return &iamdomain.InstanceProfile{
 		InstanceProfileName: "app-profile",
 		InstanceProfileID:   "AIPAEXAMPLE",
 		AccountID:           "111122223333",

spinifex/domains/ec2/guestmetadata/nats_clients.go

@@ -10,7 +10,7 @@ import (
 	"github.com/aws/aws-sdk-go/service/iam"
 	"github.com/aws/aws-sdk-go/service/sts"
 	ec2key "github.com/mulgadc/spinifex/spinifex/domains/ec2/key"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	handlers_sts "github.com/mulgadc/spinifex/spinifex/handlers/sts"
 	"github.com/nats-io/nats.go"
 )
@@ -50,7 +50,7 @@ func (a *NATSSTSAssumer) AssumeRoleForInstance(ctx context.Context, accountID, r
 // NATSProfileLookup is the NATS-backed profileLookup with a short-TTL cache.
 type NATSProfileLookup struct {
 	nc       *nats.Conn
-	profiles *ttlCache[*handlers_iam.InstanceProfile]
+	profiles *ttlCache[*iamdomain.InstanceProfile]
 	roles    *ttlCache[*iam.GetRoleOutput]
 }
 
@@ -60,17 +60,17 @@ var _ profileLookup = (*NATSProfileLookup)(nil)
 func NewNATSProfileLookup(nc *nats.Conn) *NATSProfileLookup {
 	return &NATSProfileLookup{
 		nc:       nc,
-		profiles: newTTLCache[*handlers_iam.InstanceProfile](profileCacheTTL),
+		profiles: newTTLCache[*iamdomain.InstanceProfile](profileCacheTTL),
 		roles:    newTTLCache[*iam.GetRoleOutput](profileCacheTTL),
 	}
 }
 
-func (p *NATSProfileLookup) ResolveInstanceProfile(ctx context.Context, accountID, nameOrARN string) (*handlers_iam.InstanceProfile, error) {
+func (p *NATSProfileLookup) ResolveInstanceProfile(ctx context.Context, accountID, nameOrARN string) (*iamdomain.InstanceProfile, error) {
 	key := accountID + "\x00" + nameOrARN
 	if v, ok := p.profiles.get(key); ok {
 		return v, nil
 	}
-	out, err := natsmsg.NATSRequest[handlers_iam.InstanceProfile](ctx, p.nc, handlers_iam.SubjectResolveInstanceProfile, handlers_iam.ResolveInstanceProfileRequest{
+	out, err := natsmsg.NATSRequest[iamdomain.InstanceProfile](ctx, p.nc, iamdomain.SubjectResolveInstanceProfile, iamdomain.ResolveInstanceProfileRequest{
 		AccountID: accountID,
 		NameOrARN: nameOrARN,
 	}, imdsRPCTimeout, accountID)
@@ -86,7 +86,7 @@ func (p *NATSProfileLookup) GetRole(ctx context.Context, accountID string, input
 	if v, ok := p.roles.get(key); ok {
 		return v, nil
 	}
-	out, err := natsmsg.NATSRequest[iam.GetRoleOutput](ctx, p.nc, handlers_iam.SubjectGetRole, handlers_iam.GetRoleRequest{
+	out, err := natsmsg.NATSRequest[iam.GetRoleOutput](ctx, p.nc, iamdomain.SubjectGetRole, iamdomain.GetRoleRequest{
 		AccountID: accountID,
 		Input:     input,
 	}, imdsRPCTimeout, accountID)

spinifex/domains/ec2/guestmetadata/nats_clients_test.go

@@ -13,8 +13,8 @@ import (
 	"github.com/aws/aws-sdk-go/service/sts"
 	"github.com/mulgadc/spinifex/internal/testkit"
 	ec2key "github.com/mulgadc/spinifex/spinifex/domains/ec2/key"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	handlers_sts "github.com/mulgadc/spinifex/spinifex/handlers/sts"
 	"github.com/nats-io/nats.go"
 	"github.com/stretchr/testify/assert"
@@ -72,9 +72,9 @@ func TestNATSSTSAssumer_RoundTrip(t *testing.T) {
 func TestNATSProfileLookup_CachesProfile(t *testing.T) {
 	_, nc := testutil.StartTestNATS(t)
 
-	calls := serveCounted(t, nc, handlers_iam.SubjectResolveInstanceProfile,
-		func(req *handlers_iam.ResolveInstanceProfileRequest) (*handlers_iam.InstanceProfile, error) {
-			return &handlers_iam.InstanceProfile{
+	calls := serveCounted(t, nc, iamdomain.SubjectResolveInstanceProfile,
+		func(req *iamdomain.ResolveInstanceProfileRequest) (*iamdomain.InstanceProfile, error) {
+			return &iamdomain.InstanceProfile{
 				InstanceProfileName: "app",
 				AccountID:           req.AccountID,
 				RoleName:            "app-role",
@@ -94,8 +94,8 @@ func TestNATSProfileLookup_CachesProfile(t *testing.T) {
 func TestNATSProfileLookup_CachesRole(t *testing.T) {
 	_, nc := testutil.StartTestNATS(t)
 
-	calls := serveCounted(t, nc, handlers_iam.SubjectGetRole,
-		func(req *handlers_iam.GetRoleRequest) (*iam.GetRoleOutput, error) {
+	calls := serveCounted(t, nc, iamdomain.SubjectGetRole,
+		func(req *iamdomain.GetRoleRequest) (*iam.GetRoleOutput, error) {
 			return &iam.GetRoleOutput{Role: &iam.Role{
 				RoleName: req.Input.RoleName,
 				Arn:      aws.String("arn:aws:iam::123456789012:role/app-role"),
@@ -114,8 +114,8 @@ func TestNATSProfileLookup_CachesRole(t *testing.T) {
 func TestNATSProfileLookup_PropagatesError(t *testing.T) {
 	_, nc := testutil.StartTestNATS(t)
 
-	serveCounted(t, nc, handlers_iam.SubjectResolveInstanceProfile,
-		func(_ *handlers_iam.ResolveInstanceProfileRequest) (*handlers_iam.InstanceProfile, error) {
+	serveCounted(t, nc, iamdomain.SubjectResolveInstanceProfile,
+		func(_ *iamdomain.ResolveInstanceProfileRequest) (*iamdomain.InstanceProfile, error) {
 			return nil, assert.AnError
 		})
 

spinifex/domains/ec2/guestmetadata/role_visibility_test.go

@@ -11,8 +11,8 @@ import (
 	"testing"
 	"time"
 
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
@@ -47,7 +47,7 @@ func TestProfileForReportsWhyTheRoleIsMissing(t *testing.T) {
 		{
 			name: "role resolved",
 			res:  &fakeResolver{eni: testENI(), inst: &instanceFacts{iamInstanceProfileArn: "arn:aws:iam::111122223333:instance-profile/app"}},
-			iam:  &fakeIAM{profile: &handlers_iam.InstanceProfile{ARN: "arn", InstanceProfileID: "AIPA", RoleName: "app"}},
+			iam:  &fakeIAM{profile: &iamdomain.InstanceProfile{ARN: "arn", InstanceProfileID: "AIPA", RoleName: "app"}},
 			want: roleMissNone,
 		},
 	}

spinifex/domains/ec2/guestmetadata/service.go

@@ -13,9 +13,9 @@ import (
 	"time"
 
 	"github.com/aws/aws-sdk-go/service/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvstore"
 	"github.com/mulgadc/spinifex/spinifex/foundation/telemetry"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go"
 	"github.com/nats-io/nats.go/jetstream"
 )
@@ -35,7 +35,7 @@ type listTapsFunc func(ctx context.Context) (map[string]string, error)
 
 // profileLookup is the IAMService slice needed for credential + iam/* paths.
 type profileLookup interface {
-	ResolveInstanceProfile(ctx context.Context, accountID, nameOrARN string) (*handlers_iam.InstanceProfile, error)
+	ResolveInstanceProfile(ctx context.Context, accountID, nameOrARN string) (*iamdomain.InstanceProfile, error)
 	GetRole(ctx context.Context, accountID string, input *iam.GetRoleInput) (*iam.GetRoleOutput, error)
 }
 

spinifex/domains/ecr/auth/signingkey.go

@@ -20,8 +20,8 @@ import (
 	"strings"
 	"time"
 
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvutil"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go/jetstream"
 )
 
@@ -176,7 +176,7 @@ func deleteSigningKey(ctx context.Context, kv jetstream.KeyValue, kid string) er
 // decodeSigningKey decrypts and parses one stored signing key's encrypted PKCS#8
 // PEM bytes into a SigningKey under kid.
 func decodeSigningKey(kid string, ciphertext []byte, masterKey []byte) (*SigningKey, error) {
-	pemStr, err := handlers_iam.DecryptSecret(string(ciphertext), masterKey)
+	pemStr, err := iamdomain.DecryptSecret(string(ciphertext), masterKey)
 	if err != nil {
 		return nil, fmt.Errorf("decrypt signing key %s: %w", kid, err)
 	}
@@ -203,7 +203,7 @@ func generateSigningKey(ctx context.Context, kv jetstream.KeyValue, masterKey []
 		return nil, fmt.Errorf("marshal pkcs8 private key: %w", err)
 	}
 	pemBytes := pem.EncodeToMemory(&pem.Block{Type: "PRIVATE KEY", Bytes: pkcs8})
-	ciphertext, err := handlers_iam.EncryptSecret(string(pemBytes), masterKey)
+	ciphertext, err := iamdomain.EncryptSecret(string(pemBytes), masterKey)
 	if err != nil {
 		return nil, fmt.Errorf("encrypt signing key: %w", err)
 	}

spinifex/domains/iam/account_lifecycle_test.go

@@ -1,22 +1,22 @@
-package handlers_iam_test
+package iam_test
 
 import (
 	"testing"
 
 	"github.com/mulgadc/spinifex/internal/testkit"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
 
-func newLifecycleService(t *testing.T) handlers_iam.IAMService {
+func newLifecycleService(t *testing.T) iamdomain.IAMService {
 	t.Helper()
 	_, nc, _ := testutil.StartTestJetStream(t)
 
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
 
-	svc, err := handlers_iam.NewIAMServiceImpl(t.Context(), nc, masterKey)
+	svc, err := iamdomain.NewIAMServiceImpl(t.Context(), nc, masterKey)
 	require.NoError(t, err)
 	return svc
 }
@@ -27,18 +27,18 @@ func TestSetAccountStatusMovesBetweenStates(t *testing.T) {
 	account, err := svc.CreateAccount("tenant@example.com")
 	require.NoError(t, err)
 
-	suspended, err := svc.SetAccountStatus(account.AccountID, handlers_iam.AccountStatusSuspended)
+	suspended, err := svc.SetAccountStatus(account.AccountID, iamdomain.AccountStatusSuspended)
 	require.NoError(t, err)
-	assert.Equal(t, handlers_iam.AccountStatusSuspended, suspended.Status)
+	assert.Equal(t, iamdomain.AccountStatusSuspended, suspended.Status)
 
 	// An operator hold is reversible, unlike a teardown.
-	restored, err := svc.SetAccountStatus(account.AccountID, handlers_iam.AccountStatusActive)
+	restored, err := svc.SetAccountStatus(account.AccountID, iamdomain.AccountStatusActive)
 	require.NoError(t, err)
-	assert.Equal(t, handlers_iam.AccountStatusActive, restored.Status)
+	assert.Equal(t, iamdomain.AccountStatusActive, restored.Status)
 
 	stored, err := svc.GetAccount(account.AccountID)
 	require.NoError(t, err)
-	assert.Equal(t, handlers_iam.AccountStatusActive, stored.Status)
+	assert.Equal(t, iamdomain.AccountStatusActive, stored.Status)
 }
 
 // There is no path back from TERMINATING. A half-torn-down account that
@@ -50,17 +50,17 @@ func TestTerminatingIsAOneWayDoor(t *testing.T) {
 	account, err := svc.CreateAccount("tenant@example.com")
 	require.NoError(t, err)
 
-	_, err = svc.SetAccountStatus(account.AccountID, handlers_iam.AccountStatusTerminating)
+	_, err = svc.SetAccountStatus(account.AccountID, iamdomain.AccountStatusTerminating)
 	require.NoError(t, err)
 
-	for _, status := range []string{handlers_iam.AccountStatusActive, handlers_iam.AccountStatusSuspended} {
+	for _, status := range []string{iamdomain.AccountStatusActive, iamdomain.AccountStatusSuspended} {
 		_, err = svc.SetAccountStatus(account.AccountID, status)
 		require.Error(t, err)
 	}
 
 	stored, err := svc.GetAccount(account.AccountID)
 	require.NoError(t, err)
-	assert.Equal(t, handlers_iam.AccountStatusTerminating, stored.Status)
+	assert.Equal(t, iamdomain.AccountStatusTerminating, stored.Status)
 }
 
 func TestSetAccountStatusRejectsAnUnknownStatus(t *testing.T) {
@@ -88,7 +88,7 @@ func TestDeleteAccountRequiresTerminating(t *testing.T) {
 
 	require.Error(t, svc.DeleteAccount(account.AccountID))
 
-	_, err = svc.SetAccountStatus(account.AccountID, handlers_iam.AccountStatusTerminating)
+	_, err = svc.SetAccountStatus(account.AccountID, iamdomain.AccountStatusTerminating)
 	require.NoError(t, err)
 	require.NoError(t, svc.DeleteAccount(account.AccountID))
 
@@ -102,16 +102,16 @@ func TestProtectedAccountsCanNeverBeDeleted(t *testing.T) {
 	t.Parallel()
 	svc := newLifecycleService(t)
 
-	for accountID := range handlers_iam.UndeletableAccountIDs {
+	for accountID := range iamdomain.UndeletableAccountIDs {
 		t.Run(accountID, func(t *testing.T) {
 			t.Parallel()
 			// Even a status that would otherwise permit deletion must not.
-			_, _ = svc.SetAccountStatus(accountID, handlers_iam.AccountStatusTerminating)
+			_, _ = svc.SetAccountStatus(accountID, iamdomain.AccountStatusTerminating)
 
 			err := svc.DeleteAccount(accountID)
 
 			require.Error(t, err)
-			assert.ErrorIs(t, err, handlers_iam.ErrAccountUndeletable)
+			assert.ErrorIs(t, err, iamdomain.ErrAccountUndeletable)
 		})
 	}
 }

spinifex/domains/iam/account_names.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/account_names_test.go

@@ -1,21 +1,21 @@
-package handlers_iam_test
+package iam_test
 
 import (
 	"strings"
 	"testing"
 
 	"github.com/mulgadc/spinifex/internal/testkit"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/nats-io/nats.go/jetstream"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
 
-func newAccountNameIndex(t *testing.T) (*handlers_iam.AccountNameIndex, jetstream.JetStream) {
+func newAccountNameIndex(t *testing.T) (*iamdomain.AccountNameIndex, jetstream.JetStream) {
 	t.Helper()
 	_, _, js := testutil.StartTestJetStream(t)
 
-	index, err := handlers_iam.NewAccountNameIndex(t.Context(), js)
+	index, err := iamdomain.NewAccountNameIndex(t.Context(), js)
 	require.NoError(t, err)
 	return index, js
 }
@@ -55,12 +55,12 @@ func TestAccountNameIndexReserveConflicts(t *testing.T) {
 	require.NoError(t, index.Reserve(ctx, "ben@example.com", "token-a"))
 
 	err := index.Reserve(ctx, "ben@example.com", "token-b")
-	assert.ErrorIs(t, err, handlers_iam.ErrNameInFlight)
+	assert.ErrorIs(t, err, iamdomain.ErrNameInFlight)
 
 	require.NoError(t, index.Commit(ctx, "ben@example.com", "000000000042", "token-a"))
 
 	err = index.Reserve(ctx, "ben@example.com", "token-b")
-	assert.ErrorIs(t, err, handlers_iam.ErrNameTaken)
+	assert.ErrorIs(t, err, iamdomain.ErrNameTaken)
 }
 
 // Case and whitespace must not create a second account for the same person.
@@ -77,7 +77,7 @@ func TestAccountNameIndexNormalizesName(t *testing.T) {
 	assert.True(t, found)
 	assert.Equal(t, "000000000042", accountID)
 
-	assert.ErrorIs(t, index.Reserve(ctx, "BEN@EXAMPLE.COM", "token-b"), handlers_iam.ErrNameTaken)
+	assert.ErrorIs(t, index.Reserve(ctx, "BEN@EXAMPLE.COM", "token-b"), iamdomain.ErrNameTaken)
 }
 
 func TestAccountNameIndexRelease(t *testing.T) {
@@ -89,7 +89,7 @@ func TestAccountNameIndexRelease(t *testing.T) {
 
 	// A reservation belongs to its owner; another token cannot drop it.
 	require.NoError(t, index.Release(ctx, "ben@example.com", "token-b"))
-	assert.ErrorIs(t, index.Reserve(ctx, "ben@example.com", "token-b"), handlers_iam.ErrNameInFlight)
+	assert.ErrorIs(t, index.Reserve(ctx, "ben@example.com", "token-b"), iamdomain.ErrNameInFlight)
 
 	require.NoError(t, index.Release(ctx, "ben@example.com", "token-a"))
 	require.NoError(t, index.Reserve(ctx, "ben@example.com", "token-b"))
@@ -119,7 +119,7 @@ func TestAccountNameIndexKeysCarryNoEmailAddress(t *testing.T) {
 
 	require.NoError(t, index.Reserve(ctx, "ben@example.com", "token-a"))
 
-	kv, err := js.KeyValue(ctx, handlers_iam.KVBucketAccountNames)
+	kv, err := js.KeyValue(ctx, iamdomain.KVBucketAccountNames)
 	require.NoError(t, err)
 	keys, err := kv.Keys(ctx)
 	require.NoError(t, err)
@@ -138,12 +138,12 @@ func TestFindAccountByName(t *testing.T) {
 	created, err := svc.CreateAccount("ben@example.com")
 	require.NoError(t, err)
 
-	found, err := handlers_iam.FindAccountByName(svc, "  BEN@Example.com ")
+	found, err := iamdomain.FindAccountByName(svc, "  BEN@Example.com ")
 	require.NoError(t, err)
 	require.NotNil(t, found)
 	assert.Equal(t, created.AccountID, found.AccountID)
 
-	missing, err := handlers_iam.FindAccountByName(svc, "nobody@example.com")
+	missing, err := iamdomain.FindAccountByName(svc, "nobody@example.com")
 	require.NoError(t, err)
 	assert.Nil(t, missing)
 }
@@ -157,7 +157,7 @@ func TestAccountNameIndexRejectsCorruptEntry(t *testing.T) {
 
 	require.NoError(t, index.Reserve(ctx, "ben@example.com", "token-a"))
 
-	kv, err := js.KeyValue(ctx, handlers_iam.KVBucketAccountNames)
+	kv, err := js.KeyValue(ctx, iamdomain.KVBucketAccountNames)
 	require.NoError(t, err)
 	keys, err := kv.Keys(ctx)
 	require.NoError(t, err)

spinifex/domains/iam/account_status_reader.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"
@@ -10,8 +10,8 @@ import (
 )
 
 // AccountStatusReader reads account status from spinifex-accounts and never
-// writes. It is the read-only capability EC2's settings seal uses to check
-// its TERMINATING precondition.
+// writes. It is the read-only capability EC2's settings seal and the quota
+// override seal both use to check their TERMINATING precondition.
 type AccountStatusReader struct {
 	accounts *kvstore.Bucket // AttachOnly: never creates the bucket
 }

spinifex/domains/iam/account_status_reader_test.go

@@ -1,10 +1,10 @@
-package handlers_iam_test
+package iam_test
 
 import (
 	"testing"
 
 	"github.com/mulgadc/spinifex/internal/testkit"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/nats-io/nats.go/jetstream"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -12,16 +12,16 @@ import (
 
 // newStatusReaderFixture starts a real IAM service (which creates and
 // migrates spinifex-accounts) and a reader attached to the same JetStream.
-func newStatusReaderFixture(t *testing.T) (handlers_iam.IAMService, *handlers_iam.AccountStatusReader, jetstream.JetStream) {
+func newStatusReaderFixture(t *testing.T) (iamdomain.IAMService, *iamdomain.AccountStatusReader, jetstream.JetStream) {
 	t.Helper()
 	_, nc, js := testutil.StartTestJetStream(t)
 
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
-	svc, err := handlers_iam.NewIAMServiceImpl(t.Context(), nc, masterKey)
+	svc, err := iamdomain.NewIAMServiceImpl(t.Context(), nc, masterKey)
 	require.NoError(t, err)
 
-	return svc, handlers_iam.NewAccountStatusReader(js), js
+	return svc, iamdomain.NewAccountStatusReader(js), js
 }
 
 func TestAccountStatusReader_TerminatingIsTrue(t *testing.T) {
@@ -29,7 +29,7 @@ func TestAccountStatusReader_TerminatingIsTrue(t *testing.T) {
 	svc, reader, _ := newStatusReaderFixture(t)
 	account, err := svc.CreateAccount("tenant@example.com")
 	require.NoError(t, err)
-	_, err = svc.SetAccountStatus(account.AccountID, handlers_iam.AccountStatusTerminating)
+	_, err = svc.SetAccountStatus(account.AccountID, iamdomain.AccountStatusTerminating)
 	require.NoError(t, err)
 
 	terminating, err := reader.AccountTerminating(t.Context(), account.AccountID)
@@ -48,7 +48,7 @@ func TestAccountStatusReader_ActiveAndSuspendedAreFalse(t *testing.T) {
 
 	suspended, err := svc.CreateAccount("suspended@example.com")
 	require.NoError(t, err)
-	_, err = svc.SetAccountStatus(suspended.AccountID, handlers_iam.AccountStatusSuspended)
+	_, err = svc.SetAccountStatus(suspended.AccountID, iamdomain.AccountStatusSuspended)
 	require.NoError(t, err)
 	terminating, err = reader.AccountTerminating(t.Context(), suspended.AccountID)
 	require.NoError(t, err)
@@ -68,12 +68,12 @@ func TestAccountStatusReader_MissingRecordIsFalseNotError(t *testing.T) {
 func TestAccountStatusReader_MissingBucketIsErrorAndNeverCreated(t *testing.T) {
 	t.Parallel()
 	_, nc, js := testutil.StartTestJetStream(t)
-	reader := handlers_iam.NewAccountStatusReader(js)
+	reader := iamdomain.NewAccountStatusReader(js)
 
 	_, err := reader.AccountTerminating(t.Context(), "111111111111")
 	require.Error(t, err)
 
-	_, bucketErr := js.KeyValue(t.Context(), handlers_iam.KVBucketAccounts)
+	_, bucketErr := js.KeyValue(t.Context(), iamdomain.KVBucketAccounts)
 	assert.ErrorIs(t, bucketErr, jetstream.ErrBucketNotFound)
 	_ = nc
 }
@@ -81,22 +81,22 @@ func TestAccountStatusReader_MissingBucketIsErrorAndNeverCreated(t *testing.T) {
 func TestAccountStatusReader_UndecodableValueIsError(t *testing.T) {
 	t.Parallel()
 	_, nc, js := testutil.StartTestJetStream(t)
-	_, err := handlers_iam.NewIAMServiceImpl(t.Context(), nc, mustMasterKey(t))
+	_, err := iamdomain.NewIAMServiceImpl(t.Context(), nc, mustMasterKey(t))
 	require.NoError(t, err)
 
-	kv, err := js.KeyValue(t.Context(), handlers_iam.KVBucketAccounts)
+	kv, err := js.KeyValue(t.Context(), iamdomain.KVBucketAccounts)
 	require.NoError(t, err)
 	_, err = kv.Put(t.Context(), "222222222222", []byte(`{"status":`))
 	require.NoError(t, err)
 
-	reader := handlers_iam.NewAccountStatusReader(js)
+	reader := iamdomain.NewAccountStatusReader(js)
 	_, err = reader.AccountTerminating(t.Context(), "222222222222")
 	require.Error(t, err)
 }
 
 func mustMasterKey(t *testing.T) []byte {
 	t.Helper()
-	key, err := handlers_iam.GenerateMasterKey()
+	key, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
 	return key
 }

spinifex/domains/iam/account_summary_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"testing"

spinifex/domains/iam/account_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"encoding/json"

spinifex/domains/iam/authorization.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/authorization_test.go

@@ -1,6 +1,6 @@
 //test:in-package — uses setupTestIAMService and testAccountID, unexported fixtures shared with the other handler tests.
 
-package handlers_iam
+package iam
 
 import (
 	"testing"

spinifex/domains/iam/crypto.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"crypto/rand"

spinifex/domains/iam/crypto_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"encoding/json"

spinifex/domains/iam/groups.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/groups_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"encoding/json"

spinifex/domains/iam/imds_rpc.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"github.com/aws/aws-sdk-go/service/iam"

spinifex/domains/iam/instance_profiles.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/instance_profiles_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/invariants_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"testing"

spinifex/domains/iam/managed_policies.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/managed_policies_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"errors"

spinifex/domains/iam/mock/mock.go

@@ -1,5 +1,5 @@
 // Package mock provides an in-memory implementation of
-// handlers_iam.SystemInstanceRoleEnsurer for tests.
+// iamdomain.SystemInstanceRoleEnsurer for tests.
 package mock
 
 import (
@@ -8,12 +8,12 @@ import (
 
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 )
 
 // SystemInstanceRoleEnsurer is an in-memory
-// handlers_iam.SystemInstanceRoleEnsurer. Roles and instance profiles are
+// iamdomain.SystemInstanceRoleEnsurer. Roles and instance profiles are
 // stored keyed by name; Get* miss with NoSuchEntity until the matching
 // Create* call, and a repeat Create* reports EntityAlreadyExists, mirroring
 // the live IAM API. Every method is safe for concurrent use, so callers that
@@ -59,7 +59,7 @@ type SystemInstanceRoleEnsurer struct {
 	EmptyInstanceProfileARN bool
 }
 
-var _ handlers_iam.SystemInstanceRoleEnsurer = (*SystemInstanceRoleEnsurer)(nil)
+var _ iamdomain.SystemInstanceRoleEnsurer = (*SystemInstanceRoleEnsurer)(nil)
 
 // New returns an empty SystemInstanceRoleEnsurer ready for use.
 func New() *SystemInstanceRoleEnsurer {

spinifex/domains/iam/mock/mock_test.go

@@ -6,8 +6,8 @@ import (
 
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/iam"
+	"github.com/mulgadc/spinifex/spinifex/domains/iam/mock"
 	"github.com/mulgadc/spinifex/spinifex/foundation/aws/errors"
-	"github.com/mulgadc/spinifex/spinifex/handlers/iam/mock"
 )
 
 func TestGetRole_MissingReturnsNoSuchEntity(t *testing.T) {

spinifex/domains/iam/oidc_providers.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/oidc_providers_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"net/url"

spinifex/domains/iam/policy_size_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"fmt"

spinifex/domains/iam/policy_versions.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"cmp"

spinifex/domains/iam/policy_versions_test.go

@@ -1,6 +1,6 @@
 //test:in-package — writes raw records to policiesBucket and reads resolveAttachedPolicy, both unexported.
 
-package handlers_iam
+package iam
 
 import (
 	"fmt"

spinifex/domains/iam/provision.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"errors"

spinifex/domains/iam/provision_test.go

@@ -1,4 +1,4 @@
-package handlers_iam_test
+package iam_test
 
 import (
 	"testing"
@@ -6,21 +6,21 @@ import (
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/aws/aws-sdk-go/service/iam"
 	"github.com/mulgadc/spinifex/internal/testkit"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
 
 const provisionAccountID = "000000000042"
 
-func newIAMService(t *testing.T) *handlers_iam.IAMServiceImpl {
+func newIAMService(t *testing.T) *iamdomain.IAMServiceImpl {
 	t.Helper()
 	_, nc, _ := testutil.StartTestJetStream(t)
 
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
 
-	svc, err := handlers_iam.NewIAMServiceImpl(t.Context(), nc, masterKey)
+	svc, err := iamdomain.NewIAMServiceImpl(t.Context(), nc, masterKey)
 	require.NoError(t, err)
 	return svc
 }
@@ -39,7 +39,7 @@ func TestNormalizeAccountName(t *testing.T) {
 	for _, tc := range tests {
 		t.Run(tc.name, func(t *testing.T) {
 			t.Parallel()
-			assert.Equal(t, tc.want, handlers_iam.NormalizeAccountName(tc.in))
+			assert.Equal(t, tc.want, iamdomain.NormalizeAccountName(tc.in))
 		})
 	}
 }
@@ -48,20 +48,20 @@ func TestProvisionAccountCreatesAdminIdentity(t *testing.T) {
 	t.Parallel()
 	svc := newIAMService(t)
 
-	got, err := handlers_iam.ProvisionAccount(svc, provisionAccountID, "ben@example.com")
+	got, err := iamdomain.ProvisionAccount(svc, provisionAccountID, "ben@example.com")
 	require.NoError(t, err)
 
 	assert.Equal(t, provisionAccountID, got.AccountID)
-	assert.Equal(t, handlers_iam.AdminUserName, got.AdminUser)
+	assert.Equal(t, iamdomain.AdminUserName, got.AdminUser)
 	assert.NotEmpty(t, got.AccessKeyID)
 	assert.NotEmpty(t, got.SecretAccessKey)
 
 	attached, err := svc.ListAttachedUserPolicies(provisionAccountID, &iam.ListAttachedUserPoliciesInput{
-		UserName: aws.String(handlers_iam.AdminUserName),
+		UserName: aws.String(iamdomain.AdminUserName),
 	})
 	require.NoError(t, err)
 	require.Len(t, attached.AttachedPolicies, 1)
-	assert.Equal(t, handlers_iam.AdminPolicyName, aws.StringValue(attached.AttachedPolicies[0].PolicyName))
+	assert.Equal(t, iamdomain.AdminPolicyName, aws.StringValue(attached.AttachedPolicies[0].PolicyName))
 }
 
 // A retry must finish the job rather than fail on the steps that already ran:
@@ -70,26 +70,26 @@ func TestProvisionAccountResumesPartialState(t *testing.T) {
 	t.Parallel()
 	tests := []struct {
 		name    string
-		prepare func(t *testing.T, svc handlers_iam.IAMService)
+		prepare func(t *testing.T, svc iamdomain.IAMService)
 	}{
 		{
 			name: "admin user already exists",
-			prepare: func(t *testing.T, svc handlers_iam.IAMService) {
+			prepare: func(t *testing.T, svc iamdomain.IAMService) {
 				_, err := svc.CreateUser(provisionAccountID, &iam.CreateUserInput{
-					UserName: aws.String(handlers_iam.AdminUserName),
+					UserName: aws.String(iamdomain.AdminUserName),
 				})
 				require.NoError(t, err)
 			},
 		},
 		{
 			name: "admin user and policy already exist",
-			prepare: func(t *testing.T, svc handlers_iam.IAMService) {
+			prepare: func(t *testing.T, svc iamdomain.IAMService) {
 				_, err := svc.CreateUser(provisionAccountID, &iam.CreateUserInput{
-					UserName: aws.String(handlers_iam.AdminUserName),
+					UserName: aws.String(iamdomain.AdminUserName),
 				})
 				require.NoError(t, err)
 				_, err = svc.CreatePolicy(provisionAccountID, &iam.CreatePolicyInput{
-					PolicyName:     aws.String(handlers_iam.AdminPolicyName),
+					PolicyName:     aws.String(iamdomain.AdminPolicyName),
 					PolicyDocument: aws.String(`{"Version":"2012-10-17","Statement":[{"Effect":"Allow","Action":"*","Resource":"*"}]}`),
 				})
 				require.NoError(t, err)
@@ -103,7 +103,7 @@ func TestProvisionAccountResumesPartialState(t *testing.T) {
 			svc := newIAMService(t)
 			tc.prepare(t, svc)
 
-			got, err := handlers_iam.ProvisionAccount(svc, provisionAccountID, "ben@example.com")
+			got, err := iamdomain.ProvisionAccount(svc, provisionAccountID, "ben@example.com")
 			require.NoError(t, err)
 			assert.NotEmpty(t, got.AccessKeyID)
 		})
@@ -116,15 +116,15 @@ func TestProvisionAccountLeavesExactlyOneAccessKey(t *testing.T) {
 	t.Parallel()
 	svc := newIAMService(t)
 
-	first, err := handlers_iam.ProvisionAccount(svc, provisionAccountID, "ben@example.com")
+	first, err := iamdomain.ProvisionAccount(svc, provisionAccountID, "ben@example.com")
 	require.NoError(t, err)
 
-	second, err := handlers_iam.ProvisionAccount(svc, provisionAccountID, "ben@example.com")
+	second, err := iamdomain.ProvisionAccount(svc, provisionAccountID, "ben@example.com")
 	require.NoError(t, err)
 	assert.NotEqual(t, first.AccessKeyID, second.AccessKeyID)
 
 	listed, err := svc.ListAccessKeys(provisionAccountID, &iam.ListAccessKeysInput{
-		UserName: aws.String(handlers_iam.AdminUserName),
+		UserName: aws.String(iamdomain.AdminUserName),
 	})
 	require.NoError(t, err)
 	require.Len(t, listed.AccessKeyMetadata, 1)
@@ -135,12 +135,12 @@ func TestProvisionAccountRejectsEmptyInput(t *testing.T) {
 	t.Parallel()
 	svc := newIAMService(t)
 
-	_, err := handlers_iam.ProvisionAccount(svc, "", "ben@example.com")
+	_, err := iamdomain.ProvisionAccount(svc, "", "ben@example.com")
 	assert.Error(t, err)
 
-	_, err = handlers_iam.ProvisionAccount(svc, provisionAccountID, "")
+	_, err = iamdomain.ProvisionAccount(svc, provisionAccountID, "")
 	assert.Error(t, err)
 
-	_, err = handlers_iam.ProvisionAccount(nil, provisionAccountID, "ben@example.com")
+	_, err = iamdomain.ProvisionAccount(nil, provisionAccountID, "ben@example.com")
 	assert.Error(t, err)
 }

spinifex/domains/iam/response_shapes_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"testing"

spinifex/domains/iam/retry.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/retry_test.go

@@ -1,12 +1,12 @@
-package handlers_iam_test
+package iam_test
 
 import (
 	"testing"
 	"time"
 
 	"github.com/mulgadc/spinifex/internal/testkit"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/clustersize"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/stretchr/testify/require"
 )
 
@@ -23,11 +23,11 @@ func TestNewIAMServiceWithRetry_StopsOnAConfigFault(t *testing.T) {
 	_, nc, _ := testutil.StartTestJetStream(t)
 	clustersize.RedeclareForTest(t, 0)
 
-	masterKey, err := handlers_iam.GenerateMasterKey()
+	masterKey, err := iamdomain.GenerateMasterKey()
 	require.NoError(t, err)
 
 	start := time.Now()
-	_, err = handlers_iam.NewIAMServiceWithRetry(t.Context(), nc, masterKey)
+	_, err = iamdomain.NewIAMServiceWithRetry(t.Context(), nc, masterKey)
 	elapsed := time.Since(start)
 
 	require.Error(t, err, "an undeclared cluster size must not look like a working IAM service")

spinifex/domains/iam/roles.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/roles_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"encoding/json"

spinifex/domains/iam/service.go

@@ -1,7 +1,7 @@
-// Package handlers_iam implements IAM: users, groups, roles, policies, access
+// Package iam implements IAM: users, groups, roles, policies, access
 // keys, instance profiles and OIDC providers, persisted in JetStream KV with
 // secrets sealed under the deployment master key.
-package handlers_iam
+package iam
 
 import (
 	"github.com/aws/aws-sdk-go/service/iam"

spinifex/domains/iam/service_impl.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/service_impl_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"context"

spinifex/domains/iam/system_role.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"fmt"

spinifex/domains/iam/system_role_test.go

@@ -1,13 +1,13 @@
-// package handlers_iam_test, not handlers_iam: an in-package test importing
+// package iam_test, not iam: an in-package test importing
 // the shared mock package would create an import cycle, since the mock
-// package itself imports handlers_iam for the interface compile check.
-package handlers_iam_test
+// package itself imports iam for the interface compile check.
+package iam_test
 
 import (
 	"testing"
 
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
-	iammock "github.com/mulgadc/spinifex/spinifex/handlers/iam/mock"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
+	iammock "github.com/mulgadc/spinifex/spinifex/domains/iam/mock"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
@@ -24,7 +24,7 @@ const (
 func TestEnsureSystemInstanceProfile_CreatesAll(t *testing.T) {
 	t.Parallel()
 	f := iammock.New()
-	arn, err := handlers_iam.EnsureSystemInstanceProfile(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
+	arn, err := iamdomain.EnsureSystemInstanceProfile(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
 	require.NoError(t, err)
 	assert.Equal(t, "arn:aws:iam::"+testRoleAcct+":instance-profile/"+testRoleName, arn)
 
@@ -42,10 +42,10 @@ func TestEnsureSystemInstanceProfile_CreatesAll(t *testing.T) {
 func TestEnsureSystemInstanceProfile_Idempotent(t *testing.T) {
 	t.Parallel()
 	f := iammock.New()
-	_, err := handlers_iam.EnsureSystemInstanceProfile(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
+	_, err := iamdomain.EnsureSystemInstanceProfile(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
 	require.NoError(t, err)
 
-	_, err = handlers_iam.EnsureSystemInstanceProfile(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
+	_, err = iamdomain.EnsureSystemInstanceProfile(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
 	require.NoError(t, err)
 	assert.Equal(t, 1, f.CreateRoleCalls, "role created once")
 	assert.Equal(t, 1, f.CreateInstanceProfileCalls, "profile created once")
@@ -57,11 +57,11 @@ func TestEnsureSystemInstanceProfile_Idempotent(t *testing.T) {
 func TestConvergeSystemRolePolicy_RewritesExistingRole(t *testing.T) {
 	t.Parallel()
 	f := iammock.New()
-	_, err := handlers_iam.EnsureSystemInstanceProfile(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
+	_, err := iamdomain.EnsureSystemInstanceProfile(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
 	require.NoError(t, err)
 
 	const updated = `{"Version":"2012-10-17","Statement":[{"Effect":"Allow","Action":["svc:DoThing","sts:AssumeRole"],"Resource":"*"}]}`
-	require.NoError(t, handlers_iam.ConvergeSystemRolePolicy(f, testRoleAcct, testRoleName, testPolicyName, updated))
+	require.NoError(t, iamdomain.ConvergeSystemRolePolicy(f, testRoleAcct, testRoleName, testPolicyName, updated))
 	assert.Equal(t, updated, f.RolePolicies[testRoleName])
 }
 
@@ -71,7 +71,7 @@ func TestConvergeSystemRolePolicy_RewritesExistingRole(t *testing.T) {
 func TestConvergeSystemRolePolicy_AbsentRoleIsNotProvisioned(t *testing.T) {
 	t.Parallel()
 	f := iammock.New()
-	require.NoError(t, handlers_iam.ConvergeSystemRolePolicy(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc))
+	require.NoError(t, iamdomain.ConvergeSystemRolePolicy(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc))
 
 	assert.Zero(t, f.CreateRoleCalls)
 	assert.Zero(t, f.CreateInstanceProfileCalls)
@@ -84,11 +84,11 @@ func TestConvergeSystemRolePolicy_AbsentRoleIsNotProvisioned(t *testing.T) {
 func TestConvergeSystemRolePolicy_PropagatesPutFailure(t *testing.T) {
 	t.Parallel()
 	f := iammock.New()
-	_, err := handlers_iam.EnsureSystemInstanceProfile(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
+	_, err := iamdomain.EnsureSystemInstanceProfile(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
 	require.NoError(t, err)
 
 	f.PutRolePolicyErr = assert.AnError
-	err = handlers_iam.ConvergeSystemRolePolicy(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
+	err = iamdomain.ConvergeSystemRolePolicy(f, testRoleAcct, testRoleName, testPolicyName, testPolicyDoc)
 	require.Error(t, err)
 	assert.ErrorIs(t, err, assert.AnError)
 }

spinifex/domains/iam/tagging_current_behaviour_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"testing"

spinifex/domains/iam/tags_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"fmt"

spinifex/domains/iam/types.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"encoding/json"

spinifex/domains/iam/types_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"encoding/json"

spinifex/domains/iam/user_inline_test.go

@@ -1,4 +1,4 @@
-package handlers_iam
+package iam
 
 import (
 	"encoding/json"

spinifex/domains/ochre/vector/appliance.go

@@ -11,8 +11,8 @@ import (
 	"sync"
 	"time"
 
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/mulgadc/spinifex/spinifex/foundation/state/kvstore"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
 	"github.com/nats-io/nats.go/jetstream"
 )
 
@@ -73,7 +73,7 @@ const appliancePasswordAlphabet = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrs
 
 // ApplianceRecord is the persisted singleton record for the platform Postgres
 // appliance. EncryptedPassword is AES-256-GCM ciphertext (base64, via
-// handlers_iam.EncryptSecret/DecryptSecret) — the plaintext master password
+// iamdomain.EncryptSecret/DecryptSecret) — the plaintext master password
 // is never stored, logged, or held anywhere outside a launch/Connect call.
 //
 // No exported Appliance method ever returns an ApplianceRecord: Ensure
@@ -496,7 +496,7 @@ func (a *Appliance) getRecord(ctx context.Context) (*ApplianceRecord, uint64, er
 // decryptPassword resolves rec's encrypted master password to plaintext. The
 // error path never includes the ciphertext or any decrypted material.
 func (a *Appliance) decryptPassword(rec *ApplianceRecord) (string, error) {
-	plaintext, err := handlers_iam.DecryptSecret(rec.EncryptedPassword, a.masterKey)
+	plaintext, err := iamdomain.DecryptSecret(rec.EncryptedPassword, a.masterKey)
 	if err != nil {
 		return "", fmt.Errorf("ochrevector: decrypt appliance master password: %w", err)
 	}
@@ -511,7 +511,7 @@ func newApplianceRecord(masterKey []byte) (ApplianceRecord, string, error) {
 	if err != nil {
 		return ApplianceRecord{}, "", err
 	}
-	encrypted, err := handlers_iam.EncryptSecret(password, masterKey)
+	encrypted, err := iamdomain.EncryptSecret(password, masterKey)
 	if err != nil {
 		return ApplianceRecord{}, "", fmt.Errorf("ochrevector: encrypt appliance master password: %w", err)
 	}

spinifex/domains/ochre/vector/appliance_test.go

@@ -15,7 +15,7 @@ import (
 	"time"
 
 	"github.com/mulgadc/spinifex/internal/testkit"
-	handlers_iam "github.com/mulgadc/spinifex/spinifex/handlers/iam"
+	iamdomain "github.com/mulgadc/spinifex/spinifex/domains/iam"
 	"github.com/nats-io/nats.go/jetstream"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -110,7 +110,7 @@ func TestCredentialRoundTrip(t *testing.T) {
 	require.NoError(t, err)
 	assert.Len(t, password, appliancePasswordLength)
 
-	encrypted, err := handlers_iam.EncryptSecret(password, masterKey)
+	encrypted, err := iamdomain.EncryptSecret(password, masterKey)
 	require.NoError(t, err)
 	assert.NotEqual(t, password, encrypted)
 	assert.NotContains(t, encrypted, password)
@@ -196,7 +196,7 @@ func TestEnsure_StaleProvisioningIsResumed(t *testing.T) {
 
 	seedPassword, err := generateMasterPassword()
 	require.NoError(t, err)
-	encrypted, err := handlers_iam.EncryptSecret(seedPassword, masterKey)
+	encrypted, err := iamdomain.EncryptSecret(seedPassword, masterKey)
 	require.NoError(t, err)
 
 	stale := time.Now().UTC().Add(-2 * applianceStaleAfter)
@@ -236,7 +236,7 @@ func TestEnsure_FreshProvisioningIsWaitedNotResumed(t *testing.T) {
 
 	seedPassword, err := generateMasterPassword()
 	require.NoError(t, err)
-	encrypted, err := handlers_iam.EncryptSecret(seedPassword, masterKey)
+	encrypted, err := iamdomain.EncryptSecret(seedPassword, masterKey)
 	require.NoError(t, err)
 
 	fresh := time.Now().UTC()
@@ -328,7 +328,7 @@ func TestResume_LaunchFailureLeavesRecordProvisioning(t *testing.T) {
 
 	seedPassword, err := generateMasterPassword()
 	require.NoError(t, err)
-	encrypted, err := handlers_iam.EncryptSecret(seedPassword, masterKey)
+	encrypted, err := iamdomain.EncryptSecret(seedPassword, masterKey)
 	require.NoError(t, err)
 
 	stale := time.Now().UTC().Add(-2 * applianceStaleAfter)
@@ -379,7 +379,7 @@ func TestConnect_NotAvailable(t *testing.T) {
 
 	password, err := generateMasterPassword()
 	require.NoError(t, err)
-	encrypted, err := handlers_iam.EncryptSecret(password, masterKey)
+	encrypted, err := iamdomain.EncryptSecret(password, masterKey)
 	require.NoError(t, err)
 
 	now := time.Now().UTC()
@@ -418,7 +418,7 @@ func seedAvailableAppliance(t *testing.T, appliance *Appliance, masterKey []byte
 	t.Helper()
 	password, err := generateMasterPassword()
 	require.NoError(t, err)
-	encrypted, err := handlers_iam.EncryptSecret(password, masterKey)
+	encrypted, err := iamdomain.EncryptSecret(password, masterKey)
 	require.NoError(t, err)
 
 	now := time.Now().UTC()
@@ -800,7 +800,7 @@ func TestDecryptPassword_WrongKeyErrors(t *testing.T) {
 
 	password, err := generateMasterPassword()
 	require.NoError(t, err)
-	encrypted, err := handlers_iam.EncryptSecret(password, rightKey)
+	encrypted, err := iamdomain.EncryptSecret(password, rightKey)
 	require.NoError(t, err)
 
 	appliance := &Appliance{masterKey: wrongKey}

spinifex/foundation/aws/errors/awserrors.go

@@ -658,13 +658,22 @@ func HasDependencies(kind, id string) error {
 // retryableError wraps a registered code with a suggested Retry-After
 // duration, for a 503 that a client can expect to clear on its own within a
 // bounded window (e.g. a warm-up race) rather than an open-ended outage.
-// Error() returns only the code, so code resolution is unaffected.
+// Error() returns only the code, or the wrapped error's text, so code
+// resolution is unaffected.
 type retryableError struct {
 	code       string
+	err        error
 	retryAfter time.Duration
 }
 
-func (e *retryableError) Error() string { return e.code }
+func (e *retryableError) Error() string {
+	if e.err != nil {
+		return e.err.Error()
+	}
+	return e.code
+}
+
+func (e *retryableError) Unwrap() error { return e.err }
 
 // RetryAfter wraps code (a registered ErrorLookup key) with a suggested
 // Retry-After duration for ErrorHandler to surface as a response header.
@@ -672,6 +681,13 @@ func RetryAfter(code string, d time.Duration) error {
 	return &retryableError{code: code, retryAfter: d}
 }
 
+// WithRetryAfter attaches a suggested Retry-After duration to err without
+// changing its text, unlike joining err with RetryAfter, which appends the
+// bare code on a second line.
+func WithRetryAfter(err error, d time.Duration) error {
+	return &retryableError{err: err, retryAfter: d}
+}
+
 // ResolveRetryAfter returns the Retry-After duration attached to err via
 // RetryAfter, if any, walking the same unwrap tree ResolveErrorDetail does.
 func ResolveRetryAfter(err error) (time.Duration, bool) {