The tRPC createContext was () => ({}) — the mounted dsr/consent routers
401'd every call and the dsr singleton stub threw (audit finding A11).
createTrpcContext now accepts an app resolveUser hook; web-next resolves
the session cookie through the auth feature's validateSession (denylist
included) plus a role snapshot, and threads bindProductionDsr/Consent
(or dev-seed) bindings into every request. The dsr router resolves its
binding from ctx.dsrBinding per request instead of a throwing proxy.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
bindAllProduction injected NoopRateLimit, so the sign-in budgets the
auth manifest declares were never enforced in production (audit finding
A4/B3). The production ctx now binds InMemoryRateLimit seeded from the
manifest's rateLimit budgets (manifest stays the source of truth);
dev-seed intentionally keeps Noop so local iteration never throttles.
A regression test drives sign-in through the REAL auth production
binder + app router and asserts the 6th failed attempt returns
TOO_MANY_REQUESTS while other IPs stay unaffected. In-memory counters
are per-process; multi-instance deployments need a shared IRateLimit
backend.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
clientIp was part of the public signInInputSchema, so any client could
spoof its own rate-limit bucket or dodge IP throttling entirely (audit
finding B2). The schema no longer carries it (strict parsing rejects it
with BAD_REQUEST); instead the web-next tRPC fetch adapter derives it in
createTrpcContext from x-forwarded-for (first hop) / x-real-ip — trust
caveat documented — and the router threads ctx.clientIp to the
controller as a second, server-only argument typed outside the input
schema (SignInRequestContext).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Replace the boolean bound flag with a cached promise so concurrent
callers (layout + page server components) await the same binding
operation. Prevents DI resolution before containers are populated.
Pages import feature server components and pass only route-derived
props (slug, id). All data fetching, prefetch, and hydration is owned
by the feature's .server.tsx component.
- Add USE_DEV_SEED=false branch to bindAll dispatcher
- Use dotenv-cli to inject root .env into all Turbo tasks
- Add globalDependencies for .env cache invalidation
- Add apps/web-next/middleware.ts calling withSecurityHeaders() from
core-shared/security/next; exports matcher config excluding static assets
- Update layout.tsx to call getNonce() and render <meta name="csp-nonce">
so client-side JS can read the per-request nonce
- Update instrumentation-client.ts to read nonce from csp-nonce meta tag
and pass it to initSentryClient for feedbackIntegration CSP compliance
- Add nonce option to initSentryClient (InitClientOpts.nonce) and thread
styleNonce + scriptNonce into feedbackIntegration when provided
- Add middleware test asserting all six headers, prod/dev CSP shape, and
x-nonce presence; add feedbackIntegration nonce tests to core-shared
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add withRateLimit(rateLimit, fn) in rate-limit/with-rate-limit.ts,
attaching the RateLimited brand at DI bind time
- Extend wireUseCase to accept optional rateLimit?: IRateLimit and
compose withRateLimit innermost (before analytics/audit); propagate
__rateLimited through analytics + audit inline wrappers
- Extend withSpan and withCapture PROPAGATED_BRANDS to include
__rateLimited so the outermost binding carries the brand
- Extend assertFeatureConformance to require __rateLimited brand when
manifest.useCases[name].rateLimit.length > 0; refactored into
helper functions to stay within complexity thresholds
- Add rateLimit?: IRateLimit to BindContext; default to NoopRateLimit
in web-next bindAllProduction and bindAllDevSeed aggregators
- Unit tests for withRateLimit brand attachment, factory passthrough,
and composition; synthetic fixture tests for conformance errors
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Final sweep for setup-process bookkeeping not caught by template-reset-v1.
ADRs drop Plan-N qualifiers; spec collapses the historical 11-phase
migration table; scaffolding guide drops "Phase added" column; comment
prefixes referencing R-numbers in test describes / eslint inline comments
are normalized. Architecture-level rule IDs (R40, R52, E0, J0, etc.) are
preserved where they serve as stable cross-references in ADRs.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
I3: bind-production.test.ts instrumentation orthogonality tests updated to use
bindOtelInstrumentation as primary name (bindSentryInstrumentation alias still
wired in mock setup for deprecation-alias coverage, not in assertions).
I4: as never cast in init-server-node.ts annotated with explanation of the
sdk-trace-base / sdk-node TypeScript version conflict that necessitates it.
I5: SentryLogRecordProcessor removed from @sentry/opentelemetry mock in
no-instrumentation.ts — that class does not exist in v10.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The sentry-pii-scrubber.test.ts file tested the now-deleted sentry/scrub.ts
module. PII scrubbing is now validated in core-shared/otel/pii-scrub-processor.test.ts
at the correct layer.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
apps/web-next/bind-production.ts imports bindOtelInstrumentation instead
of bindSentryInstrumentation. Test mock tracks both names via the same
spy so existing assertions on bindSentryInstrumentation still pass.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Five fixes surfaced by the branch-wide code review on the realtime layer:
- server.ts: replace dynamic `import("@repo/auth/di/container")` with a
static top-of-file import. The dynamic-import workaround from 6a0ac63 is
no longer needed once the root tsconfig + TSX_TSCONFIG_PATH expose
decorator metadata to tsx; verified by booting `pnpm dev` clean.
- server.ts: correct the inline structural type for `validateSession` to
match the real `IAuthenticationService` contract (non-nullable, throws
on invalid session) and wrap the call in try/catch so unauthenticated
bubbles to a `null` return instead of dead-code `result ? ... : null`.
- bind-production.ts: extract `maybeRegisterRealtimePing()` that wraps the
built-in ping inbound handler in the same `withSpan(withCapture(...))`
sandwich the realtime-handler generator emits (R41–R44), so the
proof-of-life channel models the convention rather than registering raw.
- bind-production.test.ts: add 4 tests for the `REALTIME_PING_DISABLED`
env-gate (registered when unset in both binders, not registered when
"true", treated as enabled when "1").
- docs/guides/realtime.md: correct the integration-test reference at
line 285 — the test does not call `bindAllDevSeed()`; it builds the
Socket.IO server inline and exercises gates 1+2 only (gates 3+4 live in
socket-io-realtime-server.test.ts).
- adr-016: add a "Known follow-ups" section recording 6 lower-priority
refinements deferred from this branch (bridge stub test scaffolding,
registry register/registerChannel precedence, channel-template dot
constraint, server bare catch{}, BindAllDeps Partial widening, AGENTS.md
anchor count phrasing).
CI gates: lint 0 errors / 4 warnings (pre-existing turbo.json warnings),
typecheck clean, 24 web-next tests pass (was 20; 4 new env-gate tests),
boundaries 0 issues across 504 files.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Extends bindAll/bindAllProduction/bindAllDevSeed to accept BindAllDeps
and thread realtime + realtimeRegistry as the 6th and 7th positional args
into each per-feature binder. Adds bindRealtimeBridge stub (empty allowlist
in v1). bindAll accepts optional deps with InMemoryRealtimeBroadcaster
fallback so existing page-level callers stay type-safe.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
End-to-end proof-of-life. bindAllDevSeed wires InMemoryEventBus +
InMemoryJobQueue across all features. signUpController publishes
auth.user.signed-up; marketing-pages' handler enqueues
marketing-pages.send-welcome-email; the in-memory queue dispatches
the wrapped job which records on RecordingMailerService.
Adds ./di/container, ./di/symbols, ./services/mailer, and
./services/recording-mailer to auth + marketing-pages exports so
the e2e test can resolve from the per-feature containers without
deep-importing.
resolveEventsAndJobsProduction wires PayloadJobsEventBus + PayloadJobQueue
against the bootstrapped Payload instance; resolveEventsAndJobsDevSeed
wires InMemoryEventBus + InMemoryJobQueue. Each per-feature binder now
receives (bus, queue) so Phase 7 generators can subscribe handlers and
register job tasks at the <gen:event-handlers>/<gen:jobs> anchors.
Adds apps/web-next/instrumentation.ts (server) and instrumentation-client.ts
(browser) hooks, wraps next.config.mjs with withSentryConfig (R52), and adds
the R38 per-app PII scrubber smoke test.
Spec deviation: extend PII_KEY_SUBSTRINGS with "ipaddress" so keys like
ipAddress trigger key-level redaction (tighter posture than the spec's
substring list; existing scrub.test.ts still passes).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Extends bind-production.test.ts with a vi.mock for @repo/core-shared/instrumentation
that wraps bindSentryInstrumentation and bindNoopInstrumentation in vi.fn (so call
counts are observable while real implementations still run). Adds 4 orthogonality
tests covering: DSN absent + production env, DSN set + dev env, Sentry + dev seed,
Noop + production — verifying instrumentation choice is independent of repo mode.
Total: 6 original + 4 new = 10 passing tests.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Introduces resolveInstrumentation() in apps/web-next/src/server/bind-production.ts:
reads WEB_NEXT_SENTRY_DSN; if set, calls bindSentryInstrumentation; else Noop.
Both bindAllProduction and bindAllDevSeed call resolveInstrumentation() at the
top. Exports __resetBindStateForTests() and __getInstrumentationForTests() for
test helpers. Adds inversify + reflect-metadata as direct dependencies of web-next.
Existing 6 tests still pass.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- CLAUDE.md Key Conventions: 'App bootstrap' rule rewritten as 'Three
binding modes per feature' — describes USE_DEV_SEED + NODE_ENV
resolution order and the new ./di/bind-dev-seed export.
- AGENTS.md (root): exports list now mentions ./ui + ./di/bind-dev-seed;
Per-feature public-API surface table gains a row; Apps section shows
the bindAll() dispatcher with three-rule logic.
- docs/architecture/vertical-feature-spec.md §6: file shape now
includes bind-dev-seed.ts, bind-dev-seed.test.ts, __seeds__/dev.ts;
package.json exports list updated to include ./di/bind-dev-seed.
- docs/architecture/data-flow-explainer.html: anatomy tree gains
__seeds__/ row; LAYERS.di description updated with new binders +
cross-link to di-explainer.html; new LAYERS.seeds entry; public-
surface card expanded to six subpaths.
- docs/superpowers/refactor-logs/2026-05-06-input-output-unification.md
§7: new 'Post-Plan-9: dev-seed binders' entry summarizing the rollout
(commits, per-feature additions, app wiring, tests, turbo, docs).
- bind-production.test.ts: dispatcher tests use vi.stubEnv (typesafe
way to test process.env in TypeScript 5+ with @types/node read-only
process.env types). 4 dispatcher tests + 2 bindAllProduction tests
= 7 tests total.
Three-rule resolution order in bindAll() (first match wins):
1. USE_DEV_SEED === 'true' → bindAllDevSeed (explicit override)
2. NODE_ENV === 'production' → bindAllProduction
3. otherwise → bindAllDevSeed (developer default)
Rationale: 'pnpm dev' should boot the app without requiring Payload to
be running locally — dev seed is the more useful default for non-
production environments. Production servers explicitly set
NODE_ENV=production and get the real binding. The USE_DEV_SEED override
remains the escape hatch (force seed in any NODE_ENV — e.g. staging
preview, design review).
bind-production.test.ts grows from 3 tests to 8 — covers the dispatcher
matrix:
- USE_DEV_SEED='true' wins even when NODE_ENV='production'
- NODE_ENV='production' (no override) → production
- NODE_ENV='development' → dev seed (default)
- NODE_ENV unset → dev seed (default)
- USE_DEV_SEED='false' treated as not-set (only the literal 'true' triggers)
- Pre-existing 'binds all five repos' test now also asserts bindProductionMedia
di-explainer.html conditions table + mode flag strings updated to match
the new three-rule logic.
apps/web-next/src/server/bind-production.ts now exports three functions:
- bindAllProduction() — production-only binders
- bindAllDevSeed() — dev-seed-only binders (NEW, calls all 5 features)
- bindAll() — dispatcher that branches on USE_DEV_SEED env var
All page/route callers (page.tsx, about/page.tsx, blog/[slug]/page.tsx,
api/trpc/[trpc]/route.ts) updated from bindAllProduction → bindAll so the
env flag actually has effect.
docs/architecture/di-explainer.html (NEW): standalone interactive page
explaining the di/ folder file-by-file, the loading sequence (8 stages),
the three binding kinds (.to / .toDynamicValue / .toConstantValue), an
interactive three-mode picker showing how the same blogContainer state
differs across default/dev-seed/production, a conditions table, and a
final card on how tests bypass DI entirely. Sister page to
data-flow-explainer.html.
Refactor log entry + canonical doc updates follow in subsequent commits.
Straggler fixes: web-next and web-tanstack app callers did not pass {}
to queryOptions()/caller calls after Plan 9 added .input(z.object({}).strict())
to siteSettings and header procedures. All 360 tests pass, full typecheck
green across 14 packages. Refactor log §7 updated with verification summary.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
