mirror of
https://github.com/shankar0123/certctl.git
synced 2026-06-07 15:32:02 +00:00
auth-bundle-1 Phase 6-7-8: bootstrap path + scope-down CLI + auditor-role split
# Phase 6 — day-0 admin bootstrap * internal/auth/bootstrap/ (new package): Strategy interface + EnvTokenStrategy with constant-time compare, one-shot consumption via sync.Mutex, optional admin-existence probe. Bundle 2's OIDC- first-admin will plug in alongside as an alternate Strategy. * BootstrapService.ValidateAndMint: validates the operator's CERTCTL_BOOTSTRAP_TOKEN, mints a 32-byte (64-hex-char) random API key value, persists the SHA-256 hash to api_keys, grants r-admin via actor_roles, AddHashed's the runtime keystore so the just- minted key authenticates the next request without restart, and records bootstrap.consume to the audit trail with category=auth. * internal/auth/keystore.go (new): KeyStore interface + StaticKeyStore (immutable env-var-only path) + MutableKeyStore (env-var keys + DB-loaded api_keys + runtime AddHashed). The auth middleware now consumes a KeyStore so the bootstrap path can extend the lookup table at runtime. * migrations/000031_api_keys.up/down.sql: api_keys table with (id, name UNIQUE, key_hash UNIQUE, tenant_id, admin, created_by, created_at, expires_at, last_used_at). Idempotent. * /v1/auth/bootstrap GET (probe) + POST (mint) — auth-exempt. Both routes documented in api/openapi.yaml + AuthExemptRouterRoutes allowlist updated. The token never leaves internal/auth/bootstrap; the minted plaintext key flows only into the HTTP response body. * Startup warning emitted when CERTCTL_BOOTSTRAP_TOKEN is set AND admin actors already exist (config drift signal). * Tests: 4 strategy invariants (empty token born disabled, wrong token=ErrInvalidToken without consumption, one-shot consumption, admin-exists closes path), 5 service tests (happy path + actor- name validation + propagation of strategy errors + nil-deps guard + 32-byte entropy budget), 8 HTTP-handler tests (status 201/410/401/400 mapping + token-leak hygiene scan of slog + audit details + Location header). Token-leak test redirects slog.Default to a buffer for the test scope. # Phase 7 — API-key migration + scope-down CLI * GET /v1/auth/keys handler + service method ListKeys backed by ActorRoleRepository.ListDistinctActors. Returns one row per (actor_id, actor_type) pair with the slice of role IDs they hold. Permission: auth.role.list. * internal/cli/auth_scope_down.go: AuthListKeys, AuthScopeDown (interactive), AuthScopeDownNonInteractive (JSON config), AuthScopeDownSuggest (--suggest with optional --apply). The synthetic actor-demo-anon is filtered out of every interactive / bulk path; non-interactive flow logs and skips it explicitly. * SuggestRoleFromAuditEvents (pure function): walks 30 days of audit events per actor and returns the narrowest matching role (admin / mcp / viewer / agent / operator) plus a one-line reason. Classification: any admin-shaped action wins; otherwise all-MCP → mcp; all-read-only → viewer; all-agent-shaped → agent; otherwise operator. Test table pins all six classifications. * CLI subcommand tree extended: 'auth keys list' + 'auth keys scope-down [--non-interactive <cfg>] [--suggest [--apply]]'. * CHANGELOG.md leads v2.1.0 with the SECURITY: AUDIT YOUR API KEYS call-out + four flow examples. # Phase 8 — auditor role + event_category column * migrations/000032_audit_category.up/down.sql: ALTER TABLE audit_events ADD COLUMN event_category TEXT NOT NULL DEFAULT 'cert_lifecycle' + CHECK constraint (cert_lifecycle/auth/config) + (event_category) and (event_category, timestamp DESC) indexes for the auditor-filter query path. WORM trigger from migration 000018 continues to enforce append-only at the DB layer (DDL is not blocked). * domain.AuditEvent gains EventCategory string (omitempty); domain.EventCategoryCertLifecycle / Auth / Config constants. * AuditService.RecordEventWithCategory sibling of RecordEvent; legacy callers stay on RecordEvent (defaults to cert_lifecycle). Auth callers (RoleService, ActorRoleService, BootstrapService) switched to RecordEventWithCategory(..., 'auth', ...). * GET /v1/audit?category=<cat>: handler accepts the optional query param, validates against the enum (400 on invalid value), dispatches through ListAuditEventsByCategory. OpenAPI updated with the new query param + AuditEvent.event_category schema. * Postgres AuditRepository.Create now writes event_category; AuditRepository.List filters on it; AuditFilter.EventCategory gates the WHERE clause. * Tests: 5 audit-category-filter HTTP tests (dispatch routing, back-compat fallback, 400 for invalid values, all 3 enum values accepted, page+category combine, JSON output surfaces the field). 3 auditor-role invariants (auditor holds exactly audit.read+audit.export, no mutating perms, disjoint from viewer except audit.read). # Cross-phase wiring * HandlerRegistry.Bootstrap field added; cmd/server/main.go wires the bootstrap service ahead of RegisterHandlers (extracted assembleNamedAPIKeys helper into auth_backfill.go, moved the keystore + bootstrap construction up alongside the auth repos). * AuthCheckResolver / AuthActorRoleService extended with ListKeys to satisfy the Phase 7 surface; existing fakes updated. * fakeAudit + mockAuditService stubs in tests gain RecordEventWithCategory + ListAuditEventsByCategory; existing tests untouched. # Verifications * gofmt -l: clean across every modified file. * go vet ./...: clean. * staticcheck across internal/auth + handler + router + cli + service + repository + cmd + domain: clean. * go test -short -count=1: green across every Bundle-1-touched package — internal/auth (incl. bootstrap), internal/api/handler, internal/api/router, internal/cli, internal/service/auth, internal/service, internal/domain/auth, internal/repository/postgres, cmd/server, cmd/cli, plus internal/scheduler, internal/api/middleware, cmd/agent, internal/mcp.
This commit is contained in:
@@ -0,0 +1,7 @@
|
||||
-- Bundle 1 Phase 6: drop the operator API-keys table. Down is destructive;
|
||||
-- keys minted by bootstrap will fail to authenticate after this runs.
|
||||
BEGIN;
|
||||
DROP INDEX IF EXISTS idx_api_keys_created_by;
|
||||
DROP INDEX IF EXISTS idx_api_keys_tenant_id;
|
||||
DROP TABLE IF EXISTS api_keys;
|
||||
COMMIT;
|
||||
@@ -0,0 +1,47 @@
|
||||
-- Bundle 1 Phase 6 (bootstrap path): runtime-minted operator API keys.
|
||||
--
|
||||
-- Pre-Bundle-1 the only operator API keys lived in CERTCTL_API_KEYS_NAMED
|
||||
-- (env-var config; static at boot). The bootstrap endpoint
|
||||
-- POST /v1/auth/bootstrap mints the first admin key without requiring
|
||||
-- the operator to know the env-var format up front; that key has to
|
||||
-- survive a process restart and authenticate against the auth
|
||||
-- middleware's keystore on subsequent requests, which means it lives
|
||||
-- here.
|
||||
--
|
||||
-- Storage rules: ONLY the SHA-256 hash of the key value is stored
|
||||
-- (key_hash). The plaintext key value is returned to the operator in
|
||||
-- the bootstrap HTTP response body once and never persisted. Lost?
|
||||
-- Mint a new admin key via the regular RBAC API and revoke the old
|
||||
-- one — the api_keys row is the source of truth for "this name +
|
||||
-- hash authenticates", so revoking it via the RBAC API removes the
|
||||
-- row and the next request lookup fails 401.
|
||||
--
|
||||
-- Idempotent: CREATE TABLE IF NOT EXISTS, indexes IF NOT EXISTS.
|
||||
|
||||
BEGIN;
|
||||
|
||||
CREATE TABLE IF NOT EXISTS api_keys (
|
||||
id TEXT PRIMARY KEY, -- prefix `ak-`
|
||||
name TEXT NOT NULL UNIQUE, -- operator-visible name; matches actor_roles.actor_id
|
||||
key_hash TEXT NOT NULL UNIQUE, -- SHA-256 hex of the plaintext key
|
||||
tenant_id TEXT NOT NULL DEFAULT 't-default'
|
||||
REFERENCES tenants(id) ON DELETE CASCADE,
|
||||
-- Admin is a denormalized hint replicated from the actor's
|
||||
-- standing role grant so the auth middleware can populate
|
||||
-- AdminKey context without joining actor_roles on every request.
|
||||
-- Source of truth remains actor_roles; this column is rebuilt by
|
||||
-- the boot loader from "actor holds r-admin?" queries.
|
||||
admin BOOLEAN NOT NULL DEFAULT FALSE,
|
||||
created_by TEXT NOT NULL, -- actor_id of the creator; "bootstrap" for the first one
|
||||
created_at TIMESTAMPTZ NOT NULL DEFAULT NOW(),
|
||||
-- Decoration columns for forward-compat: bundle 2 will add
|
||||
-- expiry + last_used + rotation tracking. Reserved as nullable
|
||||
-- now so the migration in Bundle 2 doesn't reshape the table.
|
||||
expires_at TIMESTAMPTZ,
|
||||
last_used_at TIMESTAMPTZ
|
||||
);
|
||||
|
||||
CREATE INDEX IF NOT EXISTS idx_api_keys_tenant_id ON api_keys(tenant_id);
|
||||
CREATE INDEX IF NOT EXISTS idx_api_keys_created_by ON api_keys(created_by);
|
||||
|
||||
COMMIT;
|
||||
@@ -0,0 +1,8 @@
|
||||
-- Bundle 1 Phase 8 down: drop the event_category column + indexes.
|
||||
-- Destructive — auditor-filter queries stop working after this runs.
|
||||
BEGIN;
|
||||
DROP INDEX IF EXISTS idx_audit_events_category_timestamp;
|
||||
DROP INDEX IF EXISTS idx_audit_events_event_category;
|
||||
ALTER TABLE audit_events DROP CONSTRAINT IF EXISTS audit_events_event_category_check;
|
||||
ALTER TABLE audit_events DROP COLUMN IF EXISTS event_category;
|
||||
COMMIT;
|
||||
@@ -0,0 +1,62 @@
|
||||
-- Bundle 1 Phase 8 — categorize audit events.
|
||||
--
|
||||
-- Why: post-Phase-1 the auditor role holds only audit.read +
|
||||
-- audit.export. Without a category column the auditor surface
|
||||
-- co-mingles cert-lifecycle events with auth-config mutations and
|
||||
-- config edits, which makes a "show me only the auth changes from
|
||||
-- last week" query impossible. Phase 8 adds the column + enum CHECK
|
||||
-- constraint + index so auditors can filter to the slice they care
|
||||
-- about.
|
||||
--
|
||||
-- Storage rules:
|
||||
--
|
||||
-- - cert_lifecycle (default): cert.issue, cert.renew, cert.revoke,
|
||||
-- cert.bulk_revoke, deployment.*, agent.heartbeat, etc.
|
||||
-- Existing rows backfill here.
|
||||
-- - auth: every auth.role.* / auth.key.* / auth.bootstrap.* event,
|
||||
-- plus the day-0 bootstrap.consume action from Phase 6.
|
||||
-- - config: issuer config edits, target config edits, settings
|
||||
-- mutations. Distinct from cert_lifecycle so a regulator can
|
||||
-- review "who changed the issuer wiring" separately from "who
|
||||
-- issued certs".
|
||||
--
|
||||
-- WORM trigger continues to enforce append-only at the DB layer
|
||||
-- (migration 000018). The ALTER TABLE itself is DDL, not DML, so
|
||||
-- it's not blocked by the trigger.
|
||||
--
|
||||
-- Idempotent: ADD COLUMN IF NOT EXISTS, ADD CONSTRAINT IF NOT EXISTS
|
||||
-- (Postgres 15+; uses DO blocks for older versions). The migration
|
||||
-- runner re-applies safely if the migration was partially completed.
|
||||
|
||||
BEGIN;
|
||||
|
||||
ALTER TABLE audit_events
|
||||
ADD COLUMN IF NOT EXISTS event_category TEXT NOT NULL DEFAULT 'cert_lifecycle';
|
||||
|
||||
-- CHECK constraint (idempotent via DO block; ADD CONSTRAINT IF NOT
|
||||
-- EXISTS is Postgres 15+ only).
|
||||
DO $$
|
||||
BEGIN
|
||||
IF NOT EXISTS (
|
||||
SELECT 1 FROM pg_constraint
|
||||
WHERE conname = 'audit_events_event_category_check'
|
||||
) THEN
|
||||
ALTER TABLE audit_events
|
||||
ADD CONSTRAINT audit_events_event_category_check
|
||||
CHECK (event_category IN ('cert_lifecycle', 'auth', 'config'));
|
||||
END IF;
|
||||
END$$;
|
||||
|
||||
-- Index for the auditor-filter query path. Single-column btree
|
||||
-- because event_category is low-cardinality (3 values today); the
|
||||
-- planner can still bitmap-scan with a small index.
|
||||
CREATE INDEX IF NOT EXISTS idx_audit_events_event_category
|
||||
ON audit_events(event_category);
|
||||
|
||||
-- Composite index for the most common auditor query: "auth events
|
||||
-- from last 7 days, newest first". The (category, timestamp DESC)
|
||||
-- shape lets the planner serve LIMIT-20 dashboards without sorting.
|
||||
CREATE INDEX IF NOT EXISTS idx_audit_events_category_timestamp
|
||||
ON audit_events(event_category, timestamp DESC);
|
||||
|
||||
COMMIT;
|
||||
Reference in New Issue
Block a user