certctl

gsadmin/certctl

Fork 0

mirror of https://github.com/shankar0123/certctl.git synced 2026-06-07 20:31:30 +00:00

Commit Graph

Author SHA1 Message Date

Author	SHA1	Message	Date
shankar0123	21aeed4f4e	legal: addlicense headers + normalize legacy variants (Phase 0 RED-4) Phase 0 closure (Path B2, post-rewrite): addlicense sweep — adds the canonical certctl LLC copyright + BUSL-1.1 SPDX header to every production Go file. Template: // Copyright 2026 certctl LLC. All rights reserved. // SPDX-License-Identifier: BUSL-1.1 Coverage: 338 / 338 production Go files (cmd/ + internal/, excluding _test.go and /testdata/). Pre-sweep coverage was 22 / 338 (6.5%); post-sweep is 338 / 338 (100%). Normalized 22 pre-existing legacy headers (`// Copyright (c) certctl` + `// SPDX-License-Identifier: BSL-1.1`) and 1 file using a `Certctl Contributors` attribution. The legacy SPDX ID `BSL-1.1` is non-standard; the official SPDX identifier for Business Source License 1.1 is `BUSL-1.1` (capital U). All 338 files now share the canonical form. Generated via: addlicense -c "certctl LLC" -y 2026 \ -f cowork/legal/copyright-header.tpl \ -ignore '/testdata/' -ignore '/_test.go' \ cmd/ internal/ Verification: find cmd internal -name '.go' -not -name '_test.go' \ -not -path '/testdata/' \ -exec grep -L '^// Copyright 2026 certctl LLC' {} \; \| wc -l Returns: 0 gofmt clean. Header additions are comments only, no compile impact. Closes: cowork/certctl-architecture-diligence-audit.html#fix-RED-4	2026-05-13 21:23:35 +00:00
shankar0123	af4fa12724	auth-bundle-1 Phase 8 follow-up: classify issuer/target audit rows + auditor end-to-end tests + gofmt drift Self-audit caught five real gaps in 3ef45e2; this commit closes them. # Phase 8 — issuer/target audit rows now classified as 'config' The Phase 8 prompt explicitly required existing config-mutation calls (issuer config, target config, etc.) to write event_category=config. The `3ef45e2` commit only migrated the auth service callers; the 6 issuer/target call-sites (internal/service/issuer.go: create/update/delete_issuer + internal/service/target.go: create/update/delete_target) still defaulted to cert_lifecycle. They now pass through RecordEventWithCategory(..., domain.EventCategoryConfig, ...) so auditors filtering /v1/audit?category=config see the slice the migration's docstring promised. # Auditor exit-criterion test Phase 8's exit criteria pin 'a user with the auditor role can list / export audit events but gets 403 on every other endpoint.' Bundle 1 unit invariants (auditor permission set, rbacGate behaviour) were in place but no end-to-end test walked the full set of admin perms with an auditor actor. internal/api/router/rbac_gate_integration_test.go gains TestRBACGate_AuditorRole_403sOnAdminRoutes (table-driven across all 5 admin perms — cert.bulk_revoke / crl.admin / scep.admin / est.admin / ca.hierarchy.manage) plus TestRBACGate_AuditorRole_PassesAuditReadGate (positive case for audit.read). # gofmt drift `3ef45e2` left two cosmetic struct-field-alignment diffs in internal/cli/auth.go and internal/api/handler/audit_handler_test.go that gofmt -l flagged. CI's gofmt step would have failed; gofmt -w applied; gofmt -l now clean across the repo. # CHANGELOG path-prefix CHANGELOG.md v2.1.0 used '/v1/auth/bootstrap' shorthand in the operator-facing flow examples. The actual route is '/api/v1/auth/bootstrap'; an operator copy-pasting the curl would 404. All five hits replaced. Verifications: gofmt clean, go vet ./internal/service/ ./internal/api/router/ clean, go test -short -count=1 green across internal/service + internal/api/router, including the 6 new auditor sub-tests (PASS).	2026-05-09 20:23:41 +00:00
shankar0123	b169f258de	auth-bundle-1 Phase 4 + 5: RBAC HTTP API + CLI surface Phase 4 (HTTP API): * internal/api/handler/auth.go: AuthHandler with 12 endpoints under /api/v1/auth/* — ListRoles, GetRole, CreateRole, UpdateRole, DeleteRole, ListPermissions, AddRolePermission, RemoveRolePermission, AssignRoleToKey, RevokeRoleFromKey, Me. callerFromRequest builds an authsvc.Caller from the Phase 3 ActorIDKey/ActorTypeKey/TenantIDKey context values. writeAuthError translates service + repository sentinels into HTTP status codes (401/403/404/409/400/500). 14 handler tests with in-memory fakes pin the HTTP shape + error mapping. * internal/api/router/router.go: HandlerRegistry gains an Auth field; 11 new routes registered. openapi_parity_test SpecParityExceptions extended with the new auth routes (OpenAPI YAML schema land in a Phase 4 follow-up commit so the schema review is its own atomic change; the route shape is fully documented inline via the Go type definitions until then). * cmd/server/main.go: wires the postgres auth repos (RoleRepository, PermissionRepository, ActorRoleRepository) + the Authorizer + RoleService/PermissionService/ActorRoleService into the new AuthHandler. Adds authPermissionCheckerAdapter to bridge the typed-string Authorizer signature to the auth.PermissionChecker interface (avoids an internal/auth → internal/service/auth import cycle). Phase 5 (CLI): * cmd/cli/main.go: adds 'auth' command dispatch with subcommands roles/permissions/keys/me. * internal/cli/auth.go: AuthMe, AuthListRoles, AuthGetRole, AuthListPermissions, AuthAssignRoleToKey, AuthRevokeRoleFromKey methods on Client. Mirrors the Phase 4 HTTP surface. Phase 3.5 (handler IsAdmin → middleware-wrapped RequirePermission) DEFERRED. Honest reasoning: (1) The 5 admin handlers (bulk_revocation, admin_crl_cache, admin_scep_intune, admin_est, intermediate_ca) currently gate via auth.IsAdmin checks INSIDE the handler bodies. Converting cleanly requires moving the gate to the router (auth.RequirePermission middleware wrap) AND removing the in-handler check AND rewriting the existing 3-test triplets per handler (M-008 pinned: _NonAdmin_Returns403 / _AdminExplicitFalse_Returns403 / _AdminPermitted_ForwardsActor) because the existing tests call the handler function directly, bypassing middleware. After conversion, those tests would pass without 403'ing because the gate moved away — the test invariants need to flow through a router-level integration setup instead. (2) Picking the right permission per handler is a security-review-worthy decision. Using existing operator-class perms (cert.revoke, issuer.edit) widens access from admin-only to operator-class; adding new admin-only perms (cert.bulk_revoke, crl.admin, scep.admin, est.admin, ca.hierarchy.manage) requires a migration 000030 plus a coordinated catalogue update in internal/domain/auth/validate.go. Both options are defensible but warrant a focused commit, not a 5-handler sweep mixed in with the API + CLI work. (3) The conversion can be done now without functional regressions IF we leave the in-handler IsAdmin checks in place AND add middleware wraps as defense-in-depth — but that's the worst of both worlds (legacy gate still blocks non-admin operators, defeating the point of RBAC; new gate adds runtime cost with no semantic change). A clean conversion needs the in-handler check removed. Concrete plan for Phase 3.5 (separate commit, next session): (a) add new admin-only perms via migration 000030 OR document the widening to operator-class; (b) wrap each of the 5 admin routes with auth.RequirePermission(checker, perm, nil) in router.go; (c) remove auth.IsAdmin checks from the 5 handler bodies; (d) move the M-008 _NonAdmin/_AdminExplicitFalse tests to router-level integration tests, keep _AdminPermitted as a direct handler test for actor-passthrough; (e) update m008_admin_gate_test.go registry to track auth.RequirePermission middleware wraps in router.go instead of auth.IsAdmin call sites in handler files. Verifications: go vet ./... clean; gofmt clean across all touched files; go test -short -count=1 across internal/auth, internal/service/auth, internal/api/handler, internal/api/router, internal/cli, cmd/server, cmd/cli all green (one transient too-many-open-files retry on internal/cli + internal/api/router; second run clean). Branch: dev/auth-bundle-1. Commit chain: `99a012e` (Phase 0 extract) -> `19497ee` (Phase 1 schema + repo) -> `bd54d5f` (Phase 2 service) -> `d473398` (Phase 3 primitive) -> THIS (Phase 4 + 5).	2026-05-09 16:43:48 +00:00

shankar0123

21aeed4f4e

legal: addlicense headers + normalize legacy variants (Phase 0 RED-4)

Phase 0 closure (Path B2, post-rewrite):

addlicense sweep — adds the canonical certctl LLC copyright + BUSL-1.1
SPDX header to every production Go file. Template:

  // Copyright 2026 certctl LLC. All rights reserved.
  // SPDX-License-Identifier: BUSL-1.1

Coverage: 338 / 338 production Go files (cmd/ + internal/, excluding
*_test.go and **/testdata/**). Pre-sweep coverage was 22 / 338 (6.5%);
post-sweep is 338 / 338 (100%).

Normalized 22 pre-existing legacy headers (`// Copyright (c) certctl`
+ `// SPDX-License-Identifier: BSL-1.1`) and 1 file using a
`Certctl Contributors` attribution. The legacy SPDX ID `BSL-1.1`
is non-standard; the official SPDX identifier for Business Source
License 1.1 is `BUSL-1.1` (capital U). All 338 files now share the
canonical form.

Generated via:
  addlicense -c "certctl LLC" -y 2026 \
    -f cowork/legal/copyright-header.tpl \
    -ignore '**/testdata/**' -ignore '**/*_test.go' \
    cmd/ internal/

Verification:
  find cmd internal -name '*.go' -not -name '*_test.go' \
    -not -path '*/testdata/*' \
    -exec grep -L '^// Copyright 2026 certctl LLC' {} \; | wc -l

  Returns: 0

gofmt clean. Header additions are comments only, no compile impact.

Closes: cowork/certctl-architecture-diligence-audit.html#fix-RED-4

2026-05-13 21:23:35 +00:00

shankar0123

af4fa12724

auth-bundle-1 Phase 8 follow-up: classify issuer/target audit rows + auditor end-to-end tests + gofmt drift

Self-audit caught five real gaps in 3ef45e2; this commit closes them.

# Phase 8 — issuer/target audit rows now classified as 'config'

The Phase 8 prompt explicitly required existing config-mutation
calls (issuer config, target config, etc.) to write
event_category=config. The 3ef45e2 commit only migrated the auth
service callers; the 6 issuer/target call-sites
(internal/service/issuer.go: create/update/delete_issuer +
internal/service/target.go: create/update/delete_target) still
defaulted to cert_lifecycle. They now pass through
RecordEventWithCategory(..., domain.EventCategoryConfig, ...) so
auditors filtering /v1/audit?category=config see the slice the
migration's docstring promised.

# Auditor exit-criterion test

Phase 8's exit criteria pin 'a user with the auditor role can list /
export audit events but gets 403 on every other endpoint.' Bundle 1
unit invariants (auditor permission set, rbacGate behaviour) were
in place but no end-to-end test walked the full set of admin perms
with an auditor actor. internal/api/router/rbac_gate_integration_test.go
gains TestRBACGate_AuditorRole_403sOnAdminRoutes (table-driven across
all 5 admin perms — cert.bulk_revoke / crl.admin / scep.admin /
est.admin / ca.hierarchy.manage) plus TestRBACGate_AuditorRole_PassesAuditReadGate
(positive case for audit.read).

# gofmt drift

3ef45e2 left two cosmetic struct-field-alignment diffs in
internal/cli/auth.go and internal/api/handler/audit_handler_test.go
that gofmt -l flagged. CI's gofmt step would have failed; gofmt -w
applied; gofmt -l now clean across the repo.

# CHANGELOG path-prefix

CHANGELOG.md v2.1.0 used '/v1/auth/bootstrap' shorthand in the
operator-facing flow examples. The actual route is
'/api/v1/auth/bootstrap'; an operator copy-pasting the curl would
404. All five hits replaced.

Verifications: gofmt clean, go vet ./internal/service/
./internal/api/router/ clean, go test -short -count=1 green across
internal/service + internal/api/router, including the 6 new
auditor sub-tests (PASS).

2026-05-09 20:23:41 +00:00

shankar0123

b169f258de

auth-bundle-1 Phase 4 + 5: RBAC HTTP API + CLI surface

Phase 4 (HTTP API):

* internal/api/handler/auth.go: AuthHandler with 12 endpoints under /api/v1/auth/* — ListRoles, GetRole, CreateRole, UpdateRole, DeleteRole, ListPermissions, AddRolePermission, RemoveRolePermission, AssignRoleToKey, RevokeRoleFromKey, Me. callerFromRequest builds an authsvc.Caller from the Phase 3 ActorIDKey/ActorTypeKey/TenantIDKey context values. writeAuthError translates service + repository sentinels into HTTP status codes (401/403/404/409/400/500). 14 handler tests with in-memory fakes pin the HTTP shape + error mapping.

* internal/api/router/router.go: HandlerRegistry gains an Auth field; 11 new routes registered. openapi_parity_test SpecParityExceptions extended with the new auth routes (OpenAPI YAML schema land in a Phase 4 follow-up commit so the schema review is its own atomic change; the route shape is fully documented inline via the Go type definitions until then).

* cmd/server/main.go: wires the postgres auth repos (RoleRepository, PermissionRepository, ActorRoleRepository) + the Authorizer + RoleService/PermissionService/ActorRoleService into the new AuthHandler. Adds authPermissionCheckerAdapter to bridge the typed-string Authorizer signature to the auth.PermissionChecker interface (avoids an internal/auth → internal/service/auth import cycle).

Phase 5 (CLI):

* cmd/cli/main.go: adds 'auth' command dispatch with subcommands roles/permissions/keys/me.

* internal/cli/auth.go: AuthMe, AuthListRoles, AuthGetRole, AuthListPermissions, AuthAssignRoleToKey, AuthRevokeRoleFromKey methods on Client. Mirrors the Phase 4 HTTP surface.

Phase 3.5 (handler IsAdmin → middleware-wrapped RequirePermission) DEFERRED. Honest reasoning:

(1) The 5 admin handlers (bulk_revocation, admin_crl_cache, admin_scep_intune, admin_est, intermediate_ca) currently gate via auth.IsAdmin checks INSIDE the handler bodies. Converting cleanly requires moving the gate to the router (auth.RequirePermission middleware wrap) AND removing the in-handler check AND rewriting the existing 3-test triplets per handler (M-008 pinned: _NonAdmin_Returns403 / _AdminExplicitFalse_Returns403 / _AdminPermitted_ForwardsActor) because the existing tests call the handler function directly, bypassing middleware. After conversion, those tests would pass without 403'ing because the gate moved away — the test invariants need to flow through a router-level integration setup instead.

(2) Picking the right permission per handler is a security-review-worthy decision. Using existing operator-class perms (cert.revoke, issuer.edit) widens access from admin-only to operator-class; adding new admin-only perms (cert.bulk_revoke, crl.admin, scep.admin, est.admin, ca.hierarchy.manage) requires a migration 000030 plus a coordinated catalogue update in internal/domain/auth/validate.go. Both options are defensible but warrant a focused commit, not a 5-handler sweep mixed in with the API + CLI work.

(3) The conversion can be done now without functional regressions IF we leave the in-handler IsAdmin checks in place AND add middleware wraps as defense-in-depth — but that's the worst of both worlds (legacy gate still blocks non-admin operators, defeating the point of RBAC; new gate adds runtime cost with no semantic change). A clean conversion needs the in-handler check removed.

Concrete plan for Phase 3.5 (separate commit, next session): (a) add new admin-only perms via migration 000030 OR document the widening to operator-class; (b) wrap each of the 5 admin routes with auth.RequirePermission(checker, perm, nil) in router.go; (c) remove auth.IsAdmin checks from the 5 handler bodies; (d) move the M-008 _NonAdmin/_AdminExplicitFalse tests to router-level integration tests, keep _AdminPermitted as a direct handler test for actor-passthrough; (e) update m008_admin_gate_test.go registry to track auth.RequirePermission middleware wraps in router.go instead of auth.IsAdmin call sites in handler files.

Verifications: go vet ./... clean; gofmt clean across all touched files; go test -short -count=1 across internal/auth, internal/service/auth, internal/api/handler, internal/api/router, internal/cli, cmd/server, cmd/cli all green (one transient too-many-open-files retry on internal/cli + internal/api/router; second run clean).

Branch: dev/auth-bundle-1. Commit chain: 99a012e (Phase 0 extract) -> 19497ee (Phase 1 schema + repo) -> bd54d5f (Phase 2 service) -> d473398 (Phase 3 primitive) -> THIS (Phase 4 + 5).

2026-05-09 16:43:48 +00:00

3 Commits