@stonyx/orm 0.3.2-alpha.8 → 0.3.2-alpha.80

package/src/orm-request.ts

@@ -1,3 +1,264 @@
+/**
+ * REST request handling and access enforcement for @stonyx/orm.
+ *
+ * ---------------------------------------------------------------------------
+ * THE `access()` CONTRACT: `access(request, { model, operation })`
+ * ---------------------------------------------------------------------------
+ * `auth()` calls your predicate with TWO arguments. The second is the access
+ * CONTEXT -- the structural facts about the request, which the framework
+ * already holds and which you should read INSTEAD of parsing anything:
+ *
+ *   context.model     The model this route was mounted for, as a model name:
+ *                     kebab-case, exactly as declared under
+ *                     `config.orm.paths.model` and keyed in the store --
+ *                     `'owner'`, `'animal'`, `'phone-number'`. NOT the
+ *                     pluralised, dasherized, mount-prefixed ROUTE name. It is
+ *                     read from the OrmRequest instance, fixed at mount time,
+ *                     and no request can influence it.
+ *
+ *   context.operation The operation being authorised. Exactly one of the four
+ *                     verbs `'read'`, `'create'`, `'update'`, `'delete'` --
+ *                     no second vocabulary ON THIS PATH, and never an HTTP
+ *                     method name like `'GET'`. These are the same four
+ *                     strings the permission-array return shape is written in
+ *                     (`['read', 'create']`), because both come from the one
+ *                     `methodAccessMap` below.
+ *
+ *                     NOT the hook vocabulary. `HookContext.operation`
+ *                     (`src/hooks.ts`, documented under "Hook Context Object"
+ *                     in the README) carries `'list' | 'get' | 'create' |
+ *                     'update' | 'delete'` on an identically-named key of an
+ *                     identically-shaped context object, and the access
+ *                     vocabulary collapses `list` and `get` into `'read'`. For
+ *                     one `GET /animals/1` a hook sees `'get'` and `access()`
+ *                     sees `'read'`, so a predicate cannot tell a collection
+ *                     read from a record read. `AccessOperation` makes
+ *                     `operation === 'get'` a compile error for a TypeScript
+ *                     consumer, because a predicate that stops matching falls
+ *                     through to the permission array -- the misreading is
+ *                     fail-open shaped.
+ *
+ *                     `undefined` when the dispatched method has no entry in
+ *                     that map. Express delivers `HEAD` to the `GET` handler,
+ *                     so this is reachable. It is left undefined rather than
+ *                     defaulted on purpose -- a fabricated `'read'` would turn
+ *                     an unclassified request into an authorised one. Treat
+ *                     `undefined` as "not classified" and deny.
+ *
+ * So a consumer writes `if (model === 'owner' && operation === 'read')`. There
+ * is no string to parse, no variant to miss, and no way to fail open through a
+ * URL shape nobody anticipated.
+ *
+ * WHAT THE CONTEXT DOES NOT TELL YOU: WHICH SURFACE. It names the model and
+ * the verb, not the route. Measured over the live router, six surfaces produce
+ * one identical context:
+ *
+ *     GET /owners                          { model: 'owner', operation: 'read' }
+ *     GET /owners/gina                     { model: 'owner', operation: 'read' }
+ *     GET /owners/gina/pets                { model: 'owner', operation: 'read' }
+ *     GET /owners/gina/relationships/pets  { model: 'owner', operation: 'read' }
+ *     GET /owners/archived                 { model: 'owner', operation: 'read' }
+ *     GET /owners/gina?include=pets        { model: 'owner', operation: 'read' }
+ *
+ * So a rule that depends on the SUB-PATH still needs `request.path` -- which is
+ * mount-relative and query-free, and is the one read of argument one the
+ * warning below sanctions. This repo's own fixture has such a rule: its
+ * `/archived` deny cannot be expressed from the context alone, and a predicate
+ * migrated to context-only would silently drop it, turning a deny into an
+ * allow. The related-resource and `?include=` surfaces serve ANOTHER model's
+ * records under `model: 'owner'`, and the context gives no signal of that
+ * (abofs/stonyx-orm#196).
+ *
+ * SUPERSEDED 2026-09-01 BY abofs/stonyx-orm#236/#237, AND KEPT FOR THE
+ * CONSTRAINT IT STATES RATHER THAN AS A DESCRIPTION OF THE CODE. The context
+ * now also carries `recordId` -- the DECODED route-parameter id, see
+ * `AccessContext.recordId` in ./types/orm-types.ts -- so the fixture's
+ * `/archived` deny IS expressible from the context alone, and the shipped
+ * sample no longer reads `request.path` at all. Retiring this wording WITH the
+ * measurement that retires it, rather than by deletion, is
+ * abofs/stonyx-orm#238.
+ *
+ * `record` IS NOT IN THIS CONTEXT, deliberately. `auth()` runs after route
+ * matching but BEFORE any handler executes (`@stonyx/rest-server`
+ * `src/request.ts:58-60`), so nothing has been fetched yet -- supplying a
+ * record would force a pre-fetch on every request, a second store hit and an
+ * ordering change in the middle of an authorization path. It is also
+ * unnecessary: the FUNCTION return shape already is the per-record hook. Return
+ * `(record) => boolean` and the handlers apply it to every record the request
+ * touches. Auth-time and record-time are separate decision points.
+ *
+ * THE SECOND ARGUMENT IS ADDITIVE. JavaScript ignores extra arguments, so an
+ * existing `access(request)` predicate keeps working exactly as before. The
+ * warning immediately below is therefore still live: `request` is still
+ * argument ONE, and reading it is still how predicates fail open.
+ *
+ * To reach ANOTHER model's predicate -- e.g. to check an animal while servicing
+ * an owners route -- use the boot-time registry:
+ *
+ *     const predicate = Orm.instance.getAccess('animal');
+ *     if (!predicate) return deny;
+ *     const verdict = predicate(request, { model: 'animal', operation: 'read' });
+ *
+ * `undefined` means NO PREDICATE COULD BE RESOLVED for that name -- which
+ * includes the case where the model has an access class that failed to load,
+ * because `setup-rest-server.ts` catches a load failure, warns, and publishes
+ * whatever partial map it had. It does NOT mean the model is unrestricted.
+ * Treat it as DENY, the same way `operation === undefined` is treated above.
+ *
+ * PASSING THE CONTEXT MAKES A MODEL-CORRECT ANSWER POSSIBLE. It does not make
+ * the answer model-correct on its own -- the resolved predicate has to READ it.
+ * Measured against an ARITY-1 predicate, on a request express dispatched to
+ * `GET /owners/angela`, asked about ANIMALS:
+ *
+ *     getAccess('animal')(ownersRequest, { model: 'animal', operation: 'read' })
+ *       ->  record => record.id !== 'angela' && record.id !== 'restricted'
+ *
+ * That is the OWNERS filter, and it returns `true` for animal 21 -- the record
+ * hidden on every animal surface. Under a mount that predicate recognises
+ * neither way it is worse: it falls through to
+ * `['read', 'create', 'update', 'delete']`, a full CRUD grant. Either way the
+ * context was supplied and the answer is not the animal answer, and it is wrong
+ * in the GRANTING direction, because that predicate is arity-1 and identifies
+ * its collection from the request. (Asserted on a live dispatch by AC9 in
+ * test/integration/orm-test.ts, against a deliberately arity-1 predicate.)
+ *
+ * This repo's own sample access class has since been MIGRATED to read the
+ * context (abofs/stonyx-orm#222), so `getAccess('animal')` here now answers
+ * with the animal filter. That is not true of a consumer tree: an arity-1
+ * predicate keeps working -- the second argument is additive -- and the caller
+ * has no supported way to tell which kind it got. The boot-time arity warning
+ * that surfaces one is abofs/stonyx-orm#221.
+ * So: pass the context, and do not treat a resolved predicate's answer as
+ * model-specific until that predicate has been migrated to read the context.
+ *
+ * ---------------------------------------------------------------------------
+ * DO NOT RECONSTRUCT THE REQUEST PATH INSIDE `access()`.
+ * ---------------------------------------------------------------------------
+ * You do not have to. `auth()` below hands your predicate the ACCESS CONTEXT as
+ * argument two, and `context.model` already names the collection -- see the
+ * contract section above. Argument ONE is still the raw transport artifact, and
+ * everything from here to the end of this banner is the record of what happened
+ * when predicates worked the collection out from it. IT IS HISTORY, NOT
+ * GUIDANCE: do not write any of it into a new predicate. Every attempt to
+ * identify the collection by parsing the request target has failed OPEN. Five
+ * distinct variants of the same three-line example have now been found, each
+ * after the previous was fixed, by five different people:
+ *
+ *   1. `request.url` is mount-relative under `RestServer.mountRoute`, so a
+ *      prefix match against it is ALWAYS false.
+ *   2. `request.originalUrl` carries the query string, so an anchored equality
+ *      check misses `/owners?filter[age]=30`.
+ *   3. The router is a bare `express()` (`caseSensitive: false`) while a
+ *      hand-written matcher is case-SENSITIVE, so `GET /OwNeRs/angela` walks
+ *      past it. Router-side: abofs/stonyx-rest-server#47.
+ *   4. Under a configured `ORM_REST_ROUTE` a hard-coded `/owners` matches
+ *      nothing -- environment-specifically, which is worse.
+ *   5. HTTP/1.1 permits an ABSOLUTE-FORM request-target. Express routes on
+ *      `parseurl(req).pathname`, but `originalUrl` is the raw target, so
+ *      `GET http://anything.example/owners/angela` reaches the handler with
+ *      `originalUrl === 'http://anything.example/owners/angela'`. A `/owners`
+ *      prefix match is false, `access()` falls through to whatever it returns
+ *      last, and the record comes back in full. It walks past a hard
+ *      `return false` deny the same way.
+ *
+ * The fix is not a sixth rule, and it is not a better string to match. It is to
+ * stop identifying the collection at all: read `context.model`. That is a claim
+ * about IDENTIFYING THE COLLECTION, not about the sample as a whole -- the
+ * `/archived` SUB-PATH rule is still a string match, and abofs/stonyx-orm#228 is
+ * a sixth spelling that gets past it.
+ *
+ * SUPERSEDED 2026-09-01 BY abofs/stonyx-orm#236/#237: the `/archived` rule is
+ * no longer a string match against the request target -- it compares the
+ * decoded `recordId` the framework supplies -- and abofs/stonyx-orm#228 is
+ * CLOSED. Retirement of this wording: abofs/stonyx-orm#238.
+ *
+ *   An intermediate revision of the sample read `request.baseUrl` -- the mount
+ *   Express ACTUALLY MATCHED. That closed all five variants (no query string,
+ *   not mount-relative, unaffected by absolute-form, already carrying the
+ *   configured `ORM_REST_ROUTE` prefix), but it was a transport artifact
+ *   standing in for a structural fact and the sample no longer does it.
+ *   `context.model` IS the structural fact, so variants 1, 2, 4 and 5 are
+ *   unconstructible against a migrated predicate rather than handled.
+ *
+ *   VARIANT 3 SURVIVES, and is deliberately not in that list. It is the general
+ *   shape "a hand-written matcher normalises differently from the router", and a
+ *   migrated predicate still runs one string comparison for any SUB-PATH rule --
+ *   in the shipped sample, the `/archived` deny. That comparison folds case but
+ *   does not decode, so `GET /owners/%61rchived` steps past it. See the
+ *   normalisation paragraph below and abofs/stonyx-orm#228.
+ *
+ *   SUPERSEDED 2026-09-01 BY abofs/stonyx-orm#236/#237. Variant 3 lived in that
+ *   one string comparison, and the comparison is gone: the sample compares the
+ *   decoded `recordId`. Left standing rather than edited because the same
+ *   "variant 3 survives" wording sits at four sites -- this header, README.md
+ *   twice, and test/sample/access/global-access.ts -- three of which SHIP, so
+ *   retiring one of four leaves the shipped copies contradicting each other.
+ *   Retiring all four WITH their measurement is abofs/stonyx-orm#238.
+ *
+ * ONE READ OF ARGUMENT ONE SURVIVES, AND IT MUST: `request.path`. It is
+ * mount-relative and query-free, and it is for rules that distinguish SUB-PATHS
+ * beneath the mount. The context names which model and which verb, NOT which
+ * route, so the sample's `/archived` deny cannot be expressed from the context
+ * alone and a context-ONLY rewrite would silently turn that deny into an allow.
+ *
+ * SUPERSEDED 2026-09-01 BY abofs/stonyx-orm#236/#237: NO read of argument one
+ * survives in the shipped sample. `recordId` names WHICH RECORD the route was
+ * addressed to, so the `/archived` deny is expressible from the context alone
+ * -- and it still must not be dropped; expressible is not optional. Retirement
+ * of this wording: abofs/stonyx-orm#238.
+ *
+ * NORMALISE THE WAY THE ROUTER DOES, AND CASE-FOLDING ALONE IS NOT THAT. The
+ * sample lower-cases before comparing, because a matcher stricter than the
+ * case-insensitive router can be stepped around. That closes the case gap only.
+ * Express sets `request.path` from the RAW, UNDECODED pathname while the router
+ * DECODES `:id`, so `GET /owners/%61rchived` reaches a `path === '/archived'`
+ * comparison as `/%61rchived` and walks past the deny. That gap is live in the
+ * sample and is tracked as abofs/stonyx-orm#228; the `.toLowerCase()` is not a
+ * complete normalisation recipe. Compare record ids at their real case.
+ *
+ * DO NOT FOLLOW THE PARAGRAPH ABOVE. SUPERSEDED 2026-09-01 BY
+ * abofs/stonyx-orm#236/#237, and flagged here rather than merely dated because
+ * it is an INSTRUCTION, not a stale observation. `.toLowerCase()` on the access
+ * path was measured WRONG IN BOTH DIRECTIONS AT ONCE: with a distinct owner
+ * seeded at `ARCHIVED`, `GET /owners/ARCHIVED` was a false DENY on the wrong
+ * record and `GET /owners/%41RCHIVED` a false ALLOW on that same record. A
+ * record id is a VALUE, not a literal route segment, and express's
+ * `case sensitive routing` governs literal segments only. Compare
+ * `context.recordId` AS IT ARRIVES: do not case-fold it, do not decode it, do
+ * not derive it from `request.path`. `AccessContext.recordId` in
+ * ./types/orm-types.ts is the contract and says "Do NOT case-fold it"; the same
+ * published tarball ships both files, and THIS paragraph is the one that is
+ * wrong. Retiring it WITH its measurement is abofs/stonyx-orm#238.
+ *
+ * `?? ''` is not a defence. It converts an absent request target into an empty
+ * string, which matches no collection, which falls through to the permission
+ * array -- a total grant. An input you cannot identify must DENY, and that
+ * applies to BOTH arguments: since #202 the guard and the read can sit on
+ * different objects, and a guard on argument two does not protect a read of
+ * argument one. The sample returns `false` for an absent `model` AND for an
+ * absent or non-string `request.path`, rather than falling through either way.
+ *
+ * SUPERSEDED 2026-09-01 BY abofs/stonyx-orm#236/#237 as to WHAT is guarded --
+ * the principle is unchanged. The sample no longer reads `request.path`, so it
+ * returns `false` for an absent `model` AND for an absent `recordId`
+ * (`undefined`, the one spelling `auth()` never produces). Retirement of this
+ * wording: abofs/stonyx-orm#238.
+ *
+ * THE REAL FIX IS abofs/stonyx-orm#202: `access()` should receive the model,
+ * the operation and the record. Prefer the array shape (`['read']`) or `false`
+ * until #202 lands; the function shape is what requires any matching at all.
+ *
+ * The enforcement gates in this file (GATE 0/1/2, the create rollback, the
+ * per-handler `isDenied` re-checks) are correct independently of that -- they
+ * enforce whatever predicate you return. The stopgap is the part where YOU have
+ * to work out which predicate to return.
+ *
+ * AND A PREDICATE IS NOT A GUARANTEE THAT A HIDDEN RECORD CANNOT BE MODIFIED.
+ * It is evaluated against the record the route is ADDRESSED TO, on that model
+ * only. A write to a DIFFERENT collection can still re-parent a hidden record
+ * and de-hide it -- abofs/stonyx-orm#207, which is blocked on #202 and #196.
+ * See `### Known limitations` in README.
+ */
 import { Request } from '@stonyx/rest-server';
 import Orm, { store, createRecord, updateRecord } from '@stonyx/orm';
 import { camelCaseToKebabCase } from '@stonyx/utils/string';
@@ -5,8 +266,10 @@ import { getPluralName } from './plural-registry.js';
 import { getBeforeHooks, getAfterHooks } from './hooks.js';
 import type { HookContext } from './hooks.js';
 import config from 'stonyx/config';
-import type { OrmRecord } from './types/orm-types.js';
-import { isOrmRecord } from './utils.js';
+import log from 'stonyx/log';
+import type { OrmRecord, AccessContext, AccessFunction, AccessMethod, AccessOperation, LinkageFilter } from './types/orm-types.js';
+import { isOrmRecord, NO_FREE_ID_ERROR } from './utils.js';
+import { interpretAccess, createLinkageFilter } from './access-verdict.js';
 
 interface OrmRequest$ extends Request {
   protocol?: string;
@@ -33,10 +296,9 @@ interface JsonApiResponse {
   included?: unknown[];
 }
 
-type AccessMethod = string | boolean | string[] | ((record: unknown) => boolean);
 type HandlerFn = (request: OrmRequest$, state: { [key: string]: unknown }) => unknown | Promise<unknown>;
 
-const methodAccessMap: { [key: string]: string } = {
+const methodAccessMap: { [key: string]: AccessOperation } = {
   GET: 'read',
   POST: 'create',
   DELETE: 'delete',
@@ -84,21 +346,128 @@ function getBaseUrl(request: OrmRequest$): string {
   return `${protocol}://${host}`;
 }
 
+/**
+ * The ONE coercion from a caller-supplied id to the key the store holds it
+ * under. Every id-bearing surface in this file goes through it, and none has a
+ * copy: `getId()` (URL params), `normalizeBodyId()` (JSON body), and the
+ * post-create `context.record` lookup in `_withHooks`.
+ *
+ * WHY IT IS SHARED RATHER THAN DUPLICATED. `getId()` and `normalizeBodyId()`
+ * each had their own arithmetic, and they disagreed: `parseInt(id)` versus
+ * `parseInt(id, 10)`. On a hex-shaped id that is a two-record difference --
+ *
+ *   GET  /animals/0x2391          -> record 9105   (getId    -> parseInt('0x2391')     = 9105)
+ *   POST /animals {"id":"0x2391"} -> lookup under 0 (normalize -> parseInt('0x2391',10) = 0)
+ *                                 -> a MISS, so the duplicate check was skipped and
+ *                                    createRecord OVERWROTE 9105 in place, answering 200
+ *
+ * -- a narrower form of the raw-versus-normalised divergence that the body-id
+ * normalisation was added to close, reintroduced by the fix for it. Two
+ * coercions that must agree cannot be kept in agreement by review; they have to
+ * be one function. Pinned by assertion 43.
+ *
+ * The third copy was found later and in a quieter place: `_withHooks` populated
+ * `context.record` for `create` with `isNaN(id) ? id : parseInt(id)` -- this
+ * function's body, inlined verbatim, feeding `store.get`. It was equivalent on
+ * every input reachable there, which is exactly what the two that DID diverge
+ * looked like until someone tried a hex id.
+ *
+ * SHARING IT IS NOT THE SAME AS IT BEING RIGHT EVERYWHERE. On a model declaring
+ * `id = attr('string')` a numeric-looking id is filed under the STRING key, so
+ * this coercion resolves `'9107'` to `9107` and the post-create lookup misses:
+ * `context.record` is `undefined` for an after-`create` hook. Inherited -- the
+ * inlined copy computed the same thing -- and NOT fixed here, because picking
+ * the right coercion needs the model's declared id type, which is the same
+ * structural information abofs/stonyx-orm#202 is about. Filed as
+ * abofs/stonyx-orm#209 and pinned by assertion 50, so closing it turns a test
+ * red rather than passing silently.
+ *
+ * `parseInt` and not `Number`, deliberately, and the anchor is NOT `getId`.
+ * It is `src/transforms.ts:7` -- `number: (value) => parseInt(value as string)`,
+ * also radix-less -- because that transform is what actually produces the store
+ * KEY a record is filed under. `getId` merely agrees with it. They differ from
+ * `Number` on `'1e3'` (1 vs 1000) and `'9105.5'` (9105 vs 9105.5), so switching
+ * this function to `Number` would make the lookup key disagree with the landing
+ * key on those shapes.
+ *
+ * THE CHANGE THAT WOULD BREAK THIS: adding a radix to `transforms.number`
+ * (`parseInt(value, 10)`). That is a one-word edit in a file with no connection
+ * to authorization, it would silently reopen the hex divergence in the other
+ * direction, and this comment would still read as correct. Assertion 45 pins
+ * the transform's radix-less shape directly, so that edit turns a test red
+ * rather than shipping.
+ *
+ * The reason `parseInt` is safe here is the `isNaN` gate in front of it:
+ * `parseInt('9105h')` is `9105`, and truncating a partially-valid id into a
+ * DIFFERENT VALID key is exactly how a collision lookup gets skipped. The gate
+ * rejects it as a string instead, so nothing is ever truncated. That gate, not
+ * the parser, is the load-bearing half -- assertion 43 pins it.
+ */
+function coerceId(id: string): string | number {
+  if (isNaN(id as unknown as number)) return id;
+
+  return parseInt(id);
+}
+
 function getId(params: { id?: string; [key: string]: unknown }): string | number {
   const id = params.id;
   if (!id) return '';
-  if (isNaN(id as unknown as number)) return id;
 
-  return parseInt(id);
+  return coerceId(id);
+}
+
+/**
+ * Normalise a caller-supplied BODY id to the key the store will hold it under.
+ *
+ * `getId()` above is params-shaped: it takes `{ id?: string }` off the URL,
+ * where the value is always a string and a falsy one means "no id". A JSON body
+ * id is neither -- it can arrive as a number, and `0` is a legitimate id that
+ * `getId()` would flatten to `''`.
+ *
+ * WHY THIS EXISTS AT ALL. createHandler used to look the collision up with the
+ * RAW body value while every other surface normalised through `getId()`. The
+ * store is a Map keyed by the coerced value, so `store.find(model, '21')` misses
+ * the entry held under `21` and the duplicate check is skipped by typing the id
+ * as a string. On `dev` that silently overwrote the colliding record and
+ * answered 200; combined with the denied-create rollback added for #190 it
+ * became an unauthenticated DELETE of any id. Normalising here is half of that
+ * fix -- see the rollback in createHandler for the other half.
+ *
+ * It shares `coerceId` with `getId` so the two surfaces cannot drift apart
+ * again, and differs from `getId` in exactly ONE place, below.
+ */
+function normalizeBodyId(id: string | number): string | number {
+  // Non-strings pass through untouched: a JSON body id can arrive as a number,
+  // and `getId`'s falsy-flatten must NOT apply to it -- `0` is a legitimate id
+  // and `getId` would turn it into `''`. (assertion 30 sweeps id `0`.)
+  if (typeof id !== 'string') return id;
+
+  // THE ONE DIVERGENCE FROM `getId`, and it mirrors rather than contradicts it:
+  // `getId` maps a falsy param to `''`, and `''` is the only string a body can
+  // carry that means "no id" -- `createRecord` treats it as absent and assigns a
+  // server id. Coercing it instead would make it address a real slot, because
+  // `parseInt('')` is `NaN` and a record CAN be held under `NaN` (a truthy but
+  // non-numeric id such as `'   '` survives `assignRecordId`'s falsy guard and
+  // then NaNs in the id transform). So `POST {"id":""}` would answer 409 against
+  // an unrelated record it never named. Pinned by assertion 44.
+  //
+  // Note what is deliberately NOT special-cased here any more: whitespace.
+  // `id.trim() === ''` used to short-circuit `'   '` as well, which made the
+  // body surface DISAGREE with the URL surface -- `getId({id:'   '})` is `NaN`,
+  // so `'   '` addresses the NaN slot on every other route while the collision
+  // lookup missed it. Same class of bug as the hex divergence above.
+  if (id === '') return id;
+
+  return coerceId(id);
 }
 
 function buildResponse(
   data: unknown,
   includeParam: string | undefined,
   recordOrRecords: OrmRecord | OrmRecord[],
-  options: { links?: { [key: string]: string }; baseUrl?: string } = {}
+  options: { links?: { [key: string]: string }; baseUrl?: string; linkage?: LinkageFilter } = {}
 ): JsonApiResponse {
-  const { links, baseUrl } = options;
+  const { links, baseUrl, linkage } = options;
   const response: JsonApiResponse = { data };
 
   // Add top-level links
@@ -113,7 +482,53 @@ function buildResponse(
 
   const includedRecords = collectIncludedRecords(recordOrRecords, includes);
   if (includedRecords.length > 0) {
-    response.included = includedRecords.map(record => record.toJSON?.({ baseUrl }));
+    // LINKAGE, NOT MEMBERSHIP -- and the distinction is the whole reason this
+    // line is one story's and the line above it is another's
+    // (abofs/stonyx-orm#235 and #233 respectively).
+    //
+    //   - WHICH RESOURCES REACH THIS ARRAY is decided by
+    //     `collectIncludedRecords` on the line above. That is MEMBERSHIP, it is
+    //     #233's, and it is deliberately untouched here: a hidden owner is
+    //     still a member of `included` after this change. Pinned green by
+    //     `[GUARD] #235 X1` so that #235 cannot close #233 incidentally.
+    //   - WHAT A RECORD ALREADY IN THIS ARRAY MAY NAME in its own
+    //     `relationships.*.data` is LINKAGE -- the same question #234 answers
+    //     for the primary document -- and that is what the `linkage` option
+    //     below decides. Before it, `GET /animals/1?include=owner,owner.pets`
+    //     filtered the primary document's `owner.data` to `null` and then
+    //     handed back eight PERMITTED animals in `included` each naming
+    //     `{"type":"owner","id":"angela"}` -- angela's whole `pets` set,
+    //     `[1, 3, 7, 10, 11, 15, 17, 20]`. `included` itself is NINE
+    //     resources there: those eight animals plus the hidden owner, whose
+    //     membership is #233's and not an animal. Neither #233 nor #234
+    //     closes that.
+    //
+    // THE FILTER IS THE CALLER'S, PASSED IN, NOT BUILT HERE. Both call sites
+    // already hold one for the primary document, and sharing it is what keeps
+    // the per-type verdict cache and the per-(type, id) decision cache alive
+    // across the primary document AND the sideload -- one verdict resolution
+    // per type for the whole response, pinned by `[GUARD] #235 C1`. Building a
+    // fresh filter here would resolve the consumer's `access()` once per
+    // included record instead.
+    //
+    // `linkage` IS OPTIONAL IN THE TYPE AND IS NOT OPTIONAL IN PRACTICE.
+    // Stating it precisely because the opposite claim stood here in an earlier
+    // draft of this change: BOTH of this function's callers supply a filter
+    // (`getCollectionHandler` and `getSingleHandler`, the only two), so the
+    // `undefined` branch has no live caller in this module today. It is
+    // optional so that omitting it degrades to the PRE-#234 document rather
+    // than to a denial -- `Record.toJSON` reads an ABSENT option as "no verdict
+    // was supplied" and emits linkage in full.
+    //
+    // WHAT IT MUST NEVER BE HANDED IS A NON-FUNCTION. `toJSON` does NOT read a
+    // non-function as absent: `Object.prototype.toString.call(linkage)` must be
+    // `'[object Function]'`, and anything else -- `null`, an `AsyncFunction`,
+    // and INCLUDING the primitive `true` -- DENIES every relationship on the
+    // document and logs once. `toJSON({ linkage: true })` emits `null` linkage.
+    // So do not "simplify" this to a boolean, and do not make it default to
+    // `true`: both spellings look like "allow everything" and mean the exact
+    // opposite (abofs/stonyx-orm#224).
+    response.included = includedRecords.map(record => record.toJSON?.({ baseUrl, linkage }));
   }
 
   return response;
@@ -251,12 +666,46 @@ function createFilterPredicate(filters: Filter[]): ((record: { [key: string]: un
   });
 }
 
+/**
+ * A function-style `access` return is a per-record predicate, and it is only
+ * meaningful if every surface that can hand a record to a caller consults it.
+ * Before #190 exactly one of seven did.
+ *
+ * Enforcement is deliberately post-fetch. `access` returns an opaque JS
+ * predicate and `store.findAll(model, conditions)` accepts only an equality
+ * conditions object that the SQL drivers translate to a WHERE clause, so
+ * query-layer enforcement would require a breaking change to the published
+ * `access` contract. That belongs in #197, not in a security patch. Six of the
+ * seven surfaces fetch by primary key anyway, so this costs exactly one row.
+ */
+function isDenied(filter: unknown, record: unknown): boolean {
+  if (typeof filter !== 'function') return false;
+
+  // A predicate that throws is treated as a denial. Unguarded, a throw escapes
+  // to express's default handler, which answers 500 (with a stack trace outside
+  // NODE_ENV=production) while a missing id still answers 404 -- so a
+  // record-dependent throw re-separates "hidden" from "does not exist" and
+  // hands back the oracle this whole change exists to close.
+  try {
+    return !(filter as (record: unknown) => boolean)(record);
+  } catch (error) {
+    // Denied, but not silently. A consumer predicate that throws on every
+    // record turns the whole collection into a 404 wall, and with no
+    // diagnostic that is indistinguishable from an empty database. `stonyx/log`
+    // is the module convention (see setup-rest-server.ts); optional-call
+    // because a consumer may not have configured the log types.
+    log.error?.(`[@stonyx/orm] access filter threw for model -- denying. ${error instanceof Error ? error.message : String(error)}`);
+
+    return true;
+  }
+}
+
 export default class OrmRequest extends Request {
   model: string;
-  access: (request: unknown) => AccessMethod;
+  access: AccessFunction;
   handlers: { [key: string]: { [key: string]: HandlerFn } };
 
-  constructor({ model, access }: { model: string; access: (request: unknown) => AccessMethod }) {
+  constructor({ model, access }: { model: string; access: AccessFunction }) {
     super(...arguments as unknown as unknown[]);
 
     this.model = model;
@@ -279,29 +728,62 @@ export default class OrmRequest extends Request {
       if (queryFilterPredicate) recordsToReturn = recordsToReturn.filter(queryFilterPredicate as (record: OrmRecord) => boolean);
 
       const baseUrl = getBaseUrl(request);
-      const data = recordsToReturn.map(record => record.toJSON?.({ fields: modelFields, baseUrl }));
+
+      // ONE filter per REQUEST, not one per record: it carries the per-type
+      // verdict cache and the per-(type, id) decision cache, and both are
+      // worthless if it is rebuilt inside the map. Measured on this exact
+      // surface with no `include=`: 48 linkage entries collapse to 7 distinct
+      // (type, id) pairs.
+      const linkage = createLinkageFilter(request);
+      const data = recordsToReturn.map(record => record.toJSON?.({ fields: modelFields, baseUrl, linkage }));
 
       return buildResponse(data, request.query?.include, recordsToReturn, {
         links: { self: `${baseUrl}/${pluralizedModel}` },
-        baseUrl
+        baseUrl,
+        // THE SAME filter object the primary documents above were serialized
+        // with, deliberately: it carries the caches, and rebuilding one here
+        // would re-resolve every type (abofs/stonyx-orm#235).
+        linkage
       });
     };
 
-    const getSingleHandler: HandlerFn = async (request) => {
+    const getSingleHandler: HandlerFn = async (request, { filter }) => {
       const record = await store.find(model, getId(request.params)) as OrmRecord | undefined;
       if (!record) return 404;
+      // 404, never 403: the status for "exists but filtered out" must be
+      // identical to "does not exist", or the fix trades an authorization
+      // bypass for a narrower existence oracle.
+      if (isDenied(filter, record)) return 404;
 
       const fieldsMap = parseFields(request.query);
       const modelFields = fieldsMap.get(pluralizedModel) || fieldsMap.get(model);
 
       const baseUrl = getBaseUrl(request);
-      return buildResponse(record.toJSON?.({ fields: modelFields, baseUrl }), request.query?.include, record, {
+      const linkage = createLinkageFilter(request);
+
+      // `buildResponse` IS given the filter now (abofs/stonyx-orm#235), and it
+      // is the SAME object the primary document is serialized with -- one
+      // verdict per type for the whole response, sideload included.
+      //
+      // The boundary that remains, so the next reader does not have to derive
+      // it: this closes what a record already in `included` may NAME. WHETHER a
+      // resource appears in `included` at all is MEMBERSHIP and it is
+      // abofs/stonyx-orm#233's -- a hidden owner is still a member here.
+      // Neither question closes the other.
+      return buildResponse(record.toJSON?.({ fields: modelFields, baseUrl, linkage }), request.query?.include, record, {
         links: { self: `${baseUrl}/${pluralizedModel}/${request.params.id}` },
-        baseUrl
+        baseUrl,
+        linkage
       });
     };
 
-    const createHandler: HandlerFn = async ({ body, query }) => {
+    const createHandler: HandlerFn = async (request, { filter }) => {
+      // BOUND, not destructured (abofs/stonyx-orm#235). `HandlerFn` has always
+      // delivered the request as argument one; this handler simply discarded
+      // the binding, which is why its response document named ids every read
+      // surface withholds. `createLinkageFilter` needs the live request and
+      // there is no signature change involved in giving it one.
+      const { body, query } = request;
       const { type, id, attributes, relationships: rels } = (body?.data || {}) as {
         type?: string;
         id?: string | number;
@@ -314,14 +796,113 @@ export default class OrmRequest extends Request {
       const fieldsMap = parseFields(query);
       const modelFields = fieldsMap.get(pluralizedModel) || fieldsMap.get(model);
 
-      // Check for duplicate ID
-      if (id !== undefined && await store.find(model, id)) return 409; // Conflict
+      // GATE 0 -- the POST existence oracle.
+      //
+      // The duplicate check runs before the filter and `store.find` sees hidden
+      // records, so POST leaks existence through its STATUS. A previous revision
+      // filtered the collision status (403 when the colliding record is denied,
+      // 409 when it is visible) and that is NOT sufficient, because the status
+      // of a create is a third outcome. With a payload the caller is permitted
+      // to create -- the normative case for a per-tenant filter, and the case an
+      // attacker picks -- all three are distinguishable in ONE request per id:
+      //
+      //   POST /animals {id:22,   owner:'gina'} -> 403   a HIDDEN record has this id
+      //   POST /animals {id:9500, owner:'gina'} -> 200   this id is FREE
+      //   POST /animals {id:8,    owner:'gina'} -> 409   a VISIBLE record has this id
+      //
+      // Filtering only the collision status narrows that to callers who cannot
+      // create a record they are allowed to see. It does not close it.
+      //
+      // It cannot be closed while a caller both chooses the id and learns
+      // whether the create succeeded: a successful create must answer
+      // differently from a refused one. So when a per-record filter is in force
+      // the caller does not get to choose the id at all. The refusal is
+      // UNCONDITIONAL and happens BEFORE any store lookup, so no status, and no
+      // lookup cost, can depend on whether that id exists. 403 -- the same
+      // status as a denied create -- so the two cannot be separated either.
+      //
+      // BOTH HALVES OF THAT ARE PINNED, because both were once asserted here and
+      // pinned by nothing:
+      //
+      //   status  -- assertion 22 sweeps payload x id x id-type, plus `null`.
+      //   latency -- assertion 41 asserts NO `store.find` is issued on this
+      //              path. Moving the refusal to after a lookup and returning
+      //              the same 403 left the suite green while re-opening a
+      //              hit-versus-miss timing difference on every id-bearing POST,
+      //              which is what would turn #197 from a ~0.06ms post-fetch
+      //              residual into a live timing oracle on create.
+      //
+      // AND THE GUARANTEE IS ONLY AS WIDE AS THE CHANNELS IT COVERS. It reads on
+      // the `id` member of the resource object, so it holds only while that is
+      // the ONLY way a caller id can reach `createRecord`. It was not: the
+      // relationships loop below re-admitted one under `key === "id"` and the
+      // gate never fired. Both strips -- `attributes.id` and `relationships.id`
+      // -- are therefore part of THIS gate, not tidiness, and assertion 39 pins
+      // them. Adding a third channel without a strip re-opens the oracle.
+      //
+      // Scoped to function-style `access` because that is exactly the population
+      // the oracle exists for: with no per-record filter there are no hidden
+      // records, and 409 discloses nothing GET /:id does not already.
+      //
+      // RESIDUALS, stated rather than implied.
+      //
+      //   - a caller can still learn that a collection HAS a per-record filter
+      //     (403 rather than 409/200 for an id-bearing POST). That discloses a
+      //     configuration fact, not a record.
+      //   - this gate is about ids arriving on THIS model's create route. It
+      //     says nothing about a write to ANOTHER collection: a `POST /owners`
+      //     carrying `relationships: {pets: {data: {id: 21}}}` -- or
+      //     `attributes: {pets: [21, 22]}`, which never enters the
+      //     relationships loop at all -- re-parents hidden animal 21 onto an
+      //     owner the caller may write, which changes the very field the
+      //     animals predicate reads and DE-HIDES it. Blocking that needs animal
+      //     21 checked against the ANIMAL model's predicate while servicing an
+      //     OWNERS route, i.e. cross-model access resolution: abofs/stonyx-orm
+      //     #207, blocked on #202 (`access` receives the model structurally)
+      //     and #196 (setup-rest-server discards the model->predicate map at
+      //     boot). NOT closed here, and no comment in this file may say it is.
+      //
+      // See README `### Known limitations`.
+      if (id !== undefined) {
+        if (typeof filter === 'function') return 403; // Forbidden
+
+        // `normalizeBodyId`, not the raw value: a string-typed id misses the
+        // store's numeric key, which skipped this check entirely.
+        const existing = await store.find(model, normalizeBodyId(id));
+        if (existing) return 409; // Conflict
+      }
 
       const { id: _ignoredId, ...sanitizedAttributes } = attributes || {};
 
-      // Extract relationship IDs from JSON:API relationships object
+      // Extract relationship IDs from JSON:API relationships object.
+      //
+      // `key` comes VERBATIM from the request body, so `id` is stripped here for
+      // exactly the same reason it is stripped from `attributes` on the line
+      // above -- and it must be, or GATE 0 is walked around by moving one field:
+      //
+      //   POST /animals {"id":21, ...}                    -> 403  GATE 0 fires
+      //   POST /animals {"relationships":{"id":{"data":{"id":21}}},
+      //                  "attributes":{"owner":"gina"}}   -> 200  BYPASS
+      //
+      // Top-level `id` stayed `undefined`, so GATE 0 never fired and the
+      // collision lookup never ran; `createRecord` took its last-entry-wins
+      // branch, overwrote hidden record 21 in place and reset its `owner` to a
+      // value the caller chose -- de-hiding it permanently. That is #190 itself,
+      // on the create surface. Pinned by assertion 39.
+      //
+      // The `id` member of the resource object is now the ONLY channel a caller
+      // id can arrive on FOR THIS MODEL'S OWN CREATE ROUTE, which is what makes
+      // GATE 0's guarantee checkable rather than merely asserted. It is not a
+      // statement about the record's reachability in general -- a relationship
+      // write on another collection reaches it without ever touching this
+      // handler (abofs/stonyx-orm#207). INHERITED from `dev`, which carries this
+      // loop verbatim; the general form -- the loop accepts any key, not just
+      // `id`, so a body key that is not a declared relationship is still
+      // mass-assigned -- is abofs/stonyx-orm#204.
       if (rels) {
         for (const [key, value] of Object.entries(rels)) {
+          if (key === 'id') continue;
+
           const relData = value?.data;
           if (relData && relData.id !== undefined) {
             (sanitizedAttributes as { [key: string]: unknown })[key] = relData.id;
@@ -330,16 +911,169 @@ export default class OrmRequest extends Request {
       }
 
       const recordAttributes = id !== undefined ? { id, ...sanitizedAttributes } : sanitizedAttributes;
-      const created = createRecord(model, recordAttributes as { [key: string]: unknown }, { serialize: false, _skipAutoPersist: true });
+
+      // Slot count BEFORE the write. `createRecord` writes to the store before
+      // the predicate can run, and the rollback below must be able to prove the
+      // slot it removes is one THIS REQUEST created. Identity alone cannot
+      // prove it: when `assignRecordId` lands on an occupied id, `createRecord`
+      // mutates the existing OrmRecord IN PLACE, so `store.get(...) === record`
+      // is true for a record the request did not create. The map's size is the
+      // only O(1) signal that distinguishes an insert from an overwrite.
+      const slotsBefore = store.get(model)?.size ?? 0;
+
+      // THE ONE `createRecord` FAILURE THIS ROUTE ANSWERS RATHER THAN
+      // PROPAGATES, and it is narrow on purpose.
+      //
+      // `assignRecordId` throws when it cannot derive a free store key for a
+      // server-assigned id. Unguarded that rejection is auto-forwarded -- there
+      // is no catch here, none in @stonyx/rest-server's dispatcher
+      // (dist/request.js:41-70), and express 5 hands it to its default error
+      // handler, which serialises the STACK, with absolute install paths and the
+      // internal module graph, to an unauthenticated caller outside
+      // NODE_ENV=production. That is the hazard :553-558 already names in this
+      // file, and every sibling refusal in this handler returns an integer
+      // status instead. So this one returns 409, matching the client-duplicate
+      // refusal at :713: the caller asked for a record and the collection has no
+      // id to give it.
+      //
+      // MATCHED ON THE SHARED PREFIX, not on a literal, and NOT by catching
+      // everything: `createRecord` also throws for "ORM is not ready", a
+      // read-only view and an unregistered model store, and turning any of those
+      // into a 409 would report a configuration fault as a conflict. Anything
+      // else is re-thrown unchanged.
+      let created;
+
+      try {
+        created = createRecord(model, recordAttributes as { [key: string]: unknown }, { serialize: false, _skipAutoPersist: true });
+      } catch (error) {
+        if (!(error instanceof Error) || !error.message.startsWith(NO_FREE_ID_ERROR)) throw error;
+
+        // Not silently. A collection that can no longer assign an id is a
+        // configuration fault (a non-injective id transform), and a bare 409
+        // with no diagnostic is indistinguishable from an ordinary duplicate.
+        log.error?.(`[@stonyx/orm] ${error.message}`);
+
+        return 409; // Conflict
+      }
+
       const record = isOrmRecord(created) ? created : null;
       if (!record) return 500;
 
-      return { data: record.toJSON?.({ fields: modelFields }) };
+      const createdNewSlot = (store.get(model)?.size ?? 0) > slotsBefore;
+
+      // 403 here, NOT 404. The oracle argument does not apply to create: there
+      // is no pre-existing record whose existence could leak, the caller
+      // supplied the attributes, and 404 on a mounted collection route is
+      // indistinguishable from "model not mounted" -- a genuinely different
+      // failure a developer needs to diagnose.
+      //
+      // The rollback is not optional. createRecord writes to the store BEFORE
+      // the predicate can run, so returning 403 alone would leave the record
+      // behind: a worse bug than the bypass being fixed.
+      if (isDenied(filter, record)) {
+        // ROLL BACK BY IDENTITY, NEVER BY ID. `store.remove(model, record.id)`
+        // on its own is a write primitive keyed by a value the caller may have
+        // supplied: with the raw-id collision bypass above, a denied
+        // `POST {"id":"21"}` answered 403 and DELETED hidden record 21 -- an
+        // unauthenticated deletion primitive across the whole id space, created
+        // by adding a rollback to a lookup that could be skipped.
+        //
+        // Both conditions are required and neither implies the other:
+        //   createdNewSlot -- the store grew, so this request inserted rather
+        //                     than overwrote. SURVIVOR AS OF #203, AND THAT IS
+        //                     WHAT THIS NOTE IS FOR. It used to be killable:
+        //                     `assignRecordId` returned last-INSERTED + 1, so a
+        //                     server-assigned id could land on an occupied slot,
+        //                     `createRecord` updated in place, and removing this
+        //                     half turned access-filter-enforcement-test.ts
+        //                     assertion 31 red. #203 closed that: the
+        //                     server-assigned path now walks past occupied keys,
+        //                     so no create reaching here can overwrite. Measured
+        //                     -- delete `createdNewSlot &&` below: `dev` gives
+        //                     55 pass / 1 fail with assertion 31 RED, this tree
+        //                     gives 56 pass / 0 fail, GREEN.
+        //                     KEPT ANYWAY, AND NOT FOR THE OLD REASON. Without
+        //                     it a denied create becomes `store.remove` on a key
+        //                     the caller may have influenced, which :815-820
+        //                     records as having been an unauthenticated deletion
+        //                     primitive across the whole id space. BECOMES
+        //                     KILLABLE AGAIN the moment any caller-supplied id
+        //                     can reach `createRecord` from this handler --
+        //                     which is exactly what has-many.ts:65 and
+        //                     belongs-to.ts:45 already do for ANOTHER model's
+        //                     store (abofs/stonyx-orm#207), and what a third
+        //                     un-stripped id channel would do for this one
+        //                     (#204). Do not delete it on the strength of #203
+        //                     being closed; that is the reasoning :862-867 warns
+        //                     about, one level up.
+        //   identity       -- the slot still holds the object we just created,
+        //                     so nothing between createRecord and here replaced
+        //                     it. Deleting this half SURVIVES the suite, and it
+        //                     is kept anyway. WHY IT IS REDUNDANT: there is no
+        //                     `await` anywhere between `slotsBefore` and
+        //                     `store.remove` -- the whole window is synchronous,
+        //                     so it is atomic under Node's event loop; before-
+        //                     `create` hooks run BEFORE the handler
+        //                     (`_withHooks` runs its hook loop ahead of
+        //                     `await handler(...)`), and a consumer predicate
+        //                     inside `isDenied` runs AFTER `createdNewSlot` is
+        //                     computed and cannot flip it. That is a property of
+        //                     THIS function, not of GATE 0 -- an earlier note
+        //                     credited GATE 0, which was both wrong (a caller id
+        //                     reached createRecord through the relationships
+        //                     loop, #204) and the wrong kind of reason: a guard
+        //                     justified on code sixty lines upstream gets
+        //                     silently re-armed when that code moves.
+        //                     SO IT BECOMES REACHABLE IF AN `await` IS
+        //                     INTRODUCED HERE, which is the change a future
+        //                     editor would actually make. Stated here rather
+        //                     than by reference: `docs/` is not in `files`, so
+        //                     a pointer into it resolves to nothing for anyone
+        //                     who installed this package. README carries the
+        //                     consumer-facing half.
+        if (createdNewSlot && store.get(model, record.id as string | number) === record) {
+          store.remove(model, record.id as string | number, { _skipAutoPersist: true });
+        }
+
+        return 403;
+      }
+
+      // The filter is built HERE, per invocation, and never hoisted into the
+      // OrmRequest constructor where the other per-mount values live: a verdict
+      // cached across requests answers a second caller with the first caller's
+      // authorization (src/access-verdict.ts says so at the constructor an
+      // implementer would reach for).
+      //
+      // AND IT IS BUILT AFTER `createRecord`, AFTER THE ROLLBACK WINDOW AND
+      // AFTER `isDenied`, so the record is in its final form at the call. The
+      // filter is lazy per type and per (type, id), so it cannot observe a
+      // pre-write state even if it were built earlier.
+      //
+      // `fields` is passed here and NOT in `updateHandler`: the two handlers
+      // are asymmetric on purpose (`updateHandler` has no `fieldsMap` in
+      // scope), and a single copy-pasted wiring would drop it from one of them.
+      return { data: record.toJSON?.({ fields: modelFields, linkage: createLinkageFilter(request) }) };
     };
 
-    const updateHandler: HandlerFn = async ({ body, params }) => {
+    const updateHandler: HandlerFn = async (request, { filter }) => {
+      // Bound rather than destructured, for the reason given in
+      // `createHandler` above (abofs/stonyx-orm#235). `PATCH /animals/1`
+      // returned 200 naming angela seconds after `GET /animals/1` returned
+      // `owner.data: null` for the same record -- one HTTP verb apart.
+      const { body, params } = request;
       const found = await store.find(model, getId(params));
       if (!found || !isOrmRecord(found)) return 404;
+      // Checked BEFORE any attribute is applied. 404 rather than 403 for the
+      // same reason as GET /:id -- 403 would disclose both that the record
+      // exists and that this caller specifically is excluded.
+      //
+      // NOT redundant behind GATE 1, and not defence in depth either: GATE 1's
+      // verdict is computed BEFORE the before-hook loop runs, and a before-hook
+      // is a published extension point that can change the answer -- by
+      // mutating the record, or against a predicate that closes over
+      // per-request state. This is the only re-evaluation after that window.
+      // Pinned by assertion 32; deleting it turns a 404 into an applied update.
+      if (isDenied(filter, found)) return 404;
       const record = found;
       const { attributes, relationships: rels } = (body?.data || {}) as {
         attributes?: { [key: string]: unknown };
@@ -362,6 +1096,19 @@ export default class OrmRequest extends Request {
       if (rels) {
         const relUpdates: { [key: string]: unknown } = {};
         for (const [key, value] of Object.entries(rels)) {
+          // The same missing key filter as createHandler's, and as the
+          // attribute loop directly above -- which already had it, while this
+          // loop did not. A PATCH carrying
+          // `relationships:{"id":{"data":{"id":9101}}}` reached `updateRecord`
+          // and RE-KEYED the record: the object held under store key 9102 then
+          // reported id 9101, so a visible record claimed a hidden record's
+          // identity on every surface that reads `record.id` rather than the map
+          // key. Gated by GATE 1 on the addressed record, so it is store
+          // corruption rather than a filter bypass -- but it is the same one-line
+          // omission two handlers apart. Pinned by assertion 40; INHERITED from
+          // `dev`; abofs/stonyx-orm#204.
+          if (key === 'id') continue;
+
           const relData = value?.data;
           if (relData && relData.id !== undefined) {
             relUpdates[key] = relData.id;
@@ -372,11 +1119,41 @@ export default class OrmRequest extends Request {
         }
       }
 
-      return { data: record.toJSON?.() };
+      // No `fields` and no `baseUrl`, both unchanged: `updateHandler` has no
+      // `fieldsMap` in scope, and adding `baseUrl` would put `links` on a
+      // document that has never carried them -- an unrelated behaviour change.
+      // #224 AC6's "emits `data: []` WITH links" is a statement about the READ
+      // surfaces; on these two handlers a filtered relationship and a
+      // genuinely-empty one are both a bare `{ data }`, which is what makes
+      // them indistinguishable here too.
+      return { data: record.toJSON?.({ linkage: createLinkageFilter(request) }) };
     };
 
-    const deleteHandler: HandlerFn = ({ params }) => {
-      store.remove(model, getId(params));
+    const deleteHandler: HandlerFn = async ({ params }, { filter }) => {
+      // Coerced ONCE. `getId(params)` was evaluated twice here -- once to find
+      // the record and once to remove it -- and a coercion evaluated repeatedly
+      // is a coercion that can be edited in one place and not the other, which
+      // is the defect `coerceId` exists to prevent.
+      const recordId = getId(params);
+      const record = await store.find(model, recordId) as OrmRecord | undefined;
+
+      // BEHAVIOUR CHANGE (#190): a DELETE of a record that never existed
+      // returned 204 before this change. It now returns 404, matching the
+      // denied case below. This is deliberate and load-bearing -- if a denied
+      // delete returned 404 while a missing one returned 204, the pair would be
+      // a perfect existence oracle and the whole fix would be worthless.
+      // Returning 204 for a denied delete was rejected instead: it falsely
+      // reports success for a request that changed nothing.
+      if (!record) return 404;
+      // Re-evaluated after the before-hook loop, exactly as in updateHandler --
+      // GATE 1 decided before the hooks ran. Pinned by assertion 33; deleting it
+      // turns a 404 into a destroyed record.
+      if (isDenied(filter, record)) return 404;
+
+      // Removed by the id of the record actually fetched, not by re-deriving it
+      // from the params a second time: the record the filter tested and the
+      // record removed are then provably the same one.
+      store.remove(model, record.id as string | number, { _skipAutoPersist: true });
       return 204;
     };
 
@@ -405,9 +1182,65 @@ export default class OrmRequest extends Request {
     }
   }
 
-  // Wraps a handler with before/after hook execution
+  // Wraps a handler with before/after hook execution.
+  //
+  // ===========================================================================
+  // TWO AUTHORIZATION GATES, AND EVERY EXECUTOR SITS BEHIND ONE OF THEM (#190)
+  //
+  // The defect this function was fixed for is NOT "a delete persists past a
+  // 404". It is that _withHooks has SEVERAL executors downstream of the
+  // handler, and originally the handler's response gated none of them. Three
+  // exist today:
+  //
+  //   1. sqlDb.persist          -- issues real SQL against the backing store
+  //   2. the after-hook pipeline -- the PUBLISHED consumer extension point;
+  //                                a cascade delete, a webhook, a search-index
+  //                                purge. `context.recordId` and
+  //                                `context.oldState` are populated for it.
+  //   3. Orm.db.save()          -- a full serialize-and-write of the store
+  //
+  // Gating them one at a time is how this keeps regressing, so the rule is:
+  // compute denial ONCE at each point where it becomes knowable, and keep every
+  // executor downstream of a gate. If you add a fourth executor to this
+  // function, it goes below GATE 2 or it is a security bug.
+  //
+  // GATE 1 (pre-handler) is required because before-hooks and `context.oldState`
+  // run/are built BEFORE the handler can consult the filter. Without it a denied
+  // DELETE still handed the hidden record's full contents to consumer code.
+  // GATE 2 (post-handler) covers everything the handler's status can reach.
+  // ===========================================================================
   private _withHooks(operation: string, handler: HandlerFn): HandlerFn {
     return async (request: OrmRequest$, state: { [key: string]: unknown }) => {
+      // `|| {}` so this function behaves like the relationship routes below,
+      // which declare `state` with a `= {}` default. It is unkillable through
+      // the rest-server dispatcher, which always passes `getState(req)`; it is
+      // listed as such in the guards-redundant-by-construction table rather
+      // than left silently unkillable, and it defends the WHOLE function (the
+      // context, the snapshot and the handler call all read `callState`) rather
+      // than one destructure that the next line would throw past anyway.
+      const callState = (state || {}) as { [key: string]: unknown };
+
+      // ---------------------------------------------------------------------
+      // THE AUTHORIZATION SNAPSHOT. Read ONCE, here, before any consumer code
+      // can run.
+      //
+      // `callState` is the object `auth()` planted the filter in, and it is
+      // also handed to every before-hook as `context.state` -- a published,
+      // WRITABLE extension point. So `state.filter` is an INPUT to the
+      // authorization decision, not only an output channel, and re-reading it
+      // after the hook loop lets a consumer hook disarm the filter:
+      //
+      //   beforeHook('get', 'animal', ctx => { delete ctx.state.filter })
+      //     -> GET /animals/21 turned 404 into 200
+      //     -> GET /animals    turned 20 records into 22
+      //
+      // GATE 1 already used this snapshot, so writes held; the READ handlers
+      // re-destructured `filter` from the live bag and did not. Everything
+      // downstream now reads `filter` from here, and the handler is handed
+      // `handlerState` below -- never `callState`.
+      // ---------------------------------------------------------------------
+      const { filter } = callState as { filter?: unknown };
+
       // Build context object for hooks
       const context: HookContext = {
         model: this.model,
@@ -416,12 +1249,37 @@ export default class OrmRequest extends Request {
         params: request.params,
         body: request.body,
         query: request.query,
-        state,
+        // Deliberately the LIVE object: `redirect` and `pipe` are read back off
+        // it by @stonyx/rest-server after the handler returns, so hooks must be
+        // able to write to it. What must not happen is the authorization
+        // decision reading it back, which is what the snapshot above prevents.
+        state: callState,
       };
 
       // Capture old state for operations that modify data
       if (operation === 'update' || operation === 'delete') {
         const existingRecord = await store.find(this.model, getId(request.params)) as OrmRecord | undefined;
+
+        // GATE 1 -- pre-handler. This record fetch already happened for
+        // oldState, so the check is free.
+        //
+        // Returning here rather than letting updateHandler/deleteHandler
+        // produce the same 404 is the point: everything between here and there
+        // is an executor the caller is not authorized to reach.
+        //   - context.oldState is a deep copy of the HIDDEN RECORD'S CONTENTS.
+        //     Building it and handing it to a before-hook discloses exactly what
+        //     the filter exists to hide.
+        //   - context.recordId is populated for delete BEFORE the handler runs,
+        //     which is the same shape as the sqlDb landmine one layer up:
+        //     `afterHook('delete', ctx => cascadeDelete(ctx.recordId))` destroys
+        //     children behind a correct 404.
+        //   - a before-hook may return a value and short-circuit, which would
+        //     otherwise return a response without the filter ever executing.
+        //
+        // 404, not 403, for the same reason as getSingleHandler: the status for
+        // "exists but filtered out" must equal "does not exist".
+        if (existingRecord && isDenied(filter, existingRecord)) return 404;
+
         if (existingRecord) {
           // Deep copy the record's data to preserve old state
           context.oldState = JSON.parse(JSON.stringify(existingRecord.__data || existingRecord));
@@ -441,11 +1299,52 @@ export default class OrmRequest extends Request {
       }
 
       // Execute main handler
-      const response = await handler(request, state);
+      // The handler receives the SNAPSHOT, never the live bag. `filter` is
+      // assigned LAST so it wins over anything a before-hook wrote to
+      // `callState.filter` -- including a `delete`, which the spread would
+      // otherwise carry through as an absent key. Every other key a hook adds
+      // is still visible to the handler; only the authorization input is
+      // pinned.
+      const handlerState = { ...callState, filter };
+      const response = await handler(request, handlerState);
+
+      // Set context.record for update BEFORE persist so SQL drivers can read it
+      if (operation === 'update' && (response as JsonApiResponse)?.data) {
+        context.record = store.get(this.model, getId(request.params));
+      }
 
-      // Persist to SQL database for create/update (delete is handled by store.remove auto-persist)
+      // GATE 2 -- post-handler. A denied or failed handler returns a bare status
+      // integer, and no executor below may run for one.
+      //
+      // `>= 400` deliberately covers every failure status, not just the
+      // authorization ones: a 400 (POST with no `type`) and a 409 (duplicate id)
+      // are equally requests in which nothing happened, and a persist or a
+      // cascade hook for one of them is just as wrong.
+      // `Number.isInteger` is a TYPE guard, not a behaviour guard, and it is
+      // unkillable TODAY: the only non-integer a handler in this file can
+      // return is a `{ data, links }` object, and `{} >= 400` is `false` by JS
+      // coercion, so dropping it changes no reachable outcome. It is kept
+      // because `>=` coerces rather than rejects, and the shapes it coerces
+      // are not obvious -- `[500] >= 400` is TRUE, so a handler that one day
+      // returned an array would have every response read as a denial. Listed
+      // as an equivalent mutant rather than left to read as coverage; it
+      // becomes killable the moment a handler returns anything array-like or
+      // numeric-string-like.
+      const denied = Number.isInteger(response) && (response as number) >= 400;
+
+      // EXECUTOR 1 -- SQL persistence, for all write operations.
+      //
+      // `response` is passed to sqlDb.persist below, but it is dropped at the
+      // driver boundary: _persistDelete(modelName, context) never receives it
+      // and guards only on context.recordId -- which _withHooks set above,
+      // BEFORE the handler ran. Without this gate a correct 404 still issues
+      // DELETE FROM ... WHERE id = ? on every SQL backend.
+      //
+      // No file-backed test can observe that, because Orm.instance.sqlDb is
+      // null in file/directory mode. See the stubbed-sqlDb assertions in
+      // test/unit/access-filter-enforcement-test.ts.
       const sqlDb = Orm.instance.sqlDb;
-      if (sqlDb && (operation === 'create' || operation === 'update')) {
+      if (sqlDb && WRITE_OPERATIONS.has(operation) && !denied) {
         await sqlDb.persist(operation, this.model, context, response);
       }
 
@@ -459,22 +1358,45 @@ export default class OrmRequest extends Request {
       } else if (operation === 'create' && (response as JsonApiResponse)?.data && ((response as { data: { id?: unknown } }).data.id)) {
         // For create, get the record from store using the ID from the response
         const responseData = (response as { data: { id: string | number } }).data;
-        const recordId = isNaN(responseData.id as unknown as number) ? responseData.id : parseInt(responseData.id as string);
-        context.record = store.get(this.model, recordId);
-      } else if (operation === 'update' && (response as JsonApiResponse)?.data) {
-        context.record = store.get(this.model, getId(request.params));
+        // `normalizeBodyId`, not a copy of its body. This line WAS
+        // `isNaN(id) ? id : parseInt(id)` -- `coerceId` inlined verbatim, a
+        // third coercion feeding a store lookup, sitting under a docblock that
+        // said neither surface had a copy. Equivalent on every input that can
+        // reach this truthy-guarded branch (`21`->`21`, `'21'`->`21`,
+        // `'angela'`->`'angela'`; `''` and `0` cannot reach it), so this is a
+        // de-duplication rather than a behaviour change -- and that is the
+        // point: the two that disagreed were equivalent on every input anyone
+        // checked, too.
+        context.record = store.get(this.model, normalizeBodyId(responseData.id) as string | number);
       } else if (operation === 'delete') {
         // For delete, the record may no longer exist, but we have oldState
         context.recordId = getId(request.params);
       }
 
-      // Run after hooks sequentially
-      for (const hook of getAfterHooks(operation, this.model)) {
-        await hook(context);
+      // EXECUTOR 2 -- the after-hook pipeline. This is the published consumer
+      // extension point (`afterHook` is exported from @stonyx/orm and from
+      // ./hooks), so it is the executor with the widest possible blast radius:
+      // a cascade delete, a webhook, a token revocation, a search-index purge.
+      //
+      // BEHAVIOUR CHANGE (#190): after-hooks no longer fire for a request that
+      // failed. Previously `afterHook('delete', ...)` ran with a populated
+      // context.recordId on a 404, so a consumer cascade destroyed children for
+      // a request that deleted nothing. Firing a hook named "after<operation>"
+      // for an operation that did not occur is a booby trap, and the denied case
+      // is unreachable-before-#190 while the missing case is inherited debt --
+      // both are closed by the same gate. `context.response` therefore only ever
+      // carries a success status into a hook.
+      if (!denied) {
+        for (const hook of getAfterHooks(operation, this.model)) {
+          await hook(context);
+        }
       }
 
-      // Auto-save DB after write operations when configured
-      if (config.orm.db.autosave === 'onUpdate' && WRITE_OPERATIONS.has(operation)) {
+      // EXECUTOR 3 -- file/directory autosave. Ungated this let an
+      // unauthenticated caller force a full serialize-and-write of the entire
+      // store on every DELETE of any id, with no record touched: amplification
+      // rather than corruption, but the same root cause and the same fix.
+      if (config.orm.db.autosave === 'onUpdate' && WRITE_OPERATIONS.has(operation) && !denied) {
         await (Orm.db as { save(): Promise<void> }).save();
       }
 
@@ -494,21 +1416,38 @@ export default class OrmRequest extends Request {
       const dasherizedName = camelCaseToKebabCase(relationshipName);
 
       // Related resource route: GET /:id/{relationship}
-      routes[`/:id/${dasherizedName}`] = async (request: OrmRequest$) => {
+      //
+      // These generated routes are not wrapped by _withHooks, which is why they
+      // were the least obvious two of the seven unguarded surfaces in #190.
+      // They are still dispatched by @stonyx/rest-server as
+      // `handler(req, getState(req))`, so `state` -- and therefore the filter
+      // planted by auth() -- has always been available here; it was simply
+      // never declared or read.
+      routes[`/:id/${dasherizedName}`] = async (request: OrmRequest$, { filter }: { [key: string]: unknown } = {}) => {
         const record = await store.find(model, getId(request.params)) as OrmRecord | undefined;
         if (!record) return 404;
+        // Filtering the PARENT: a caller who may not see the record may not see
+        // what it is related to either.
+        if (isDenied(filter, record)) return 404;
 
         const relatedData = record.__relationships[relationshipName];
         const baseUrl = getBaseUrl(request);
 
+        // LINKAGE ONLY. This filter decides which ids the emitted documents may
+        // NAME in their own `relationships.*.data`; it does NOT decide whether
+        // the related records themselves are served -- that is the parent-only
+        // filtering this route has done since #190, and widening it to the
+        // related record is abofs/stonyx-orm#196.
+        const linkage = createLinkageFilter(request);
+
         let data: unknown;
         if (info.isArray) {
           // hasMany - return array
           const related = Array.isArray(relatedData) ? relatedData.filter(isOrmRecord) : [];
-          data = related.map(r => r.toJSON?.({ baseUrl }));
+          data = related.map(r => r.toJSON?.({ baseUrl, linkage }));
         } else {
           // belongsTo - return single or null
-          data = isOrmRecord(relatedData) ? relatedData.toJSON?.({ baseUrl }) : null;
+          data = isOrmRecord(relatedData) ? relatedData.toJSON?.({ baseUrl, linkage }) : null;
         }
 
         return {
@@ -518,9 +1457,58 @@ export default class OrmRequest extends Request {
       };
 
       // Relationship linkage route: GET /:id/relationships/{relationship}
-      routes[`/:id/relationships/${dasherizedName}`] = async (request: OrmRequest$) => {
+      //
+      // NO `linkage` FILTER FROM abofs/stonyx-orm#235, AND THAT IS A SCOPE
+      // BOUNDARY RATHER THAN AN OVERSIGHT -- abofs/stonyx-orm#232 OWNS THIS
+      // ROUTE, and PR #247 is IN FLIGHT against it in this same sprint. If you
+      // are reading this after #247 landed, the filtering below is #232's and
+      // this note records why it was never #235's to add.
+      //
+      // The three sites #235 does own -- `buildResponse`'s `included`, and the
+      // two write handlers, `POST /:models` and `PATCH /:models/:id` -- all
+      // reach the filter through `record.toJSON()`, which is where the
+      // `linkage` OPTION is applied.
+      //
+      // The related-resource branch above ALSO passes a `linkage` filter, and
+      // it is NOT one of those three: it is abofs/stonyx-orm#234's code and
+      // predates this change. `git diff 8dda5d6..HEAD -- src/orm-request.ts`
+      // leaves that branch byte-unchanged.
+      //
+      // This branch builds its `{ type, id }` objects BY
+      // HAND and never calls `toJSON` at all, so the `linkage` option cannot
+      // reach it -- whatever this route filters, it has to filter itself, which
+      // is precisely why doing so is a separate change with a separate owner.
+      //
+      // It is also a DIFFERENT QUESTION. Everywhere #235 touches, linkage is
+      // metadata ABOUT a document. Here the linkage IS the primary data, so
+      // dropping an entry is a MEMBERSHIP decision about what this route
+      // serves -- the same class as abofs/stonyx-orm#233 and #196, not the
+      // class #234/#235 close. That is why it is absent from #224 §2a's
+      // seven-site inventory.
+      //
+      // MEASURED, so the next person does not re-derive it. Against this
+      // branch's baseline of 1011/0, wiring `createLinkageFilter` into the
+      // belongsTo branch below takes the suite to 1009/2, reddening
+      // `[GUARD] #235 X2` and the
+      // `GET /animals/:id/relationships/owner returns relationship linkage`
+      // test -- the latter is #232's own reproduction, not a regression.
+      //
+      // THE BASELINE IS QUOTED WITH THE RESULT BECAUSE AN EARLIER REVISION OF
+      // THIS COMMENT SAID 993/2 AND SHIPPED IT. This file lands in consumers'
+      // `node_modules`, so a wrong number here is a wrong number in the
+      // published package. 993+2 = 995 is the DEV baseline, carried over from
+      // a branch on which `[GUARD] #235 X2` does not exist. A pass/fail pair
+      // with no baseline beside it cannot be checked by reading, which is how
+      // it survived three artifacts and a review; the qualitative claim was
+      // right the whole time and only the count was wrong.
+      //
+      // `[GUARD] #235 X2` in test/integration/orm-test.ts pins the OWNERSHIP
+      // BOUNDARY here rather than this route's current answer, so that it
+      // survives #247 landing. Read its comment before changing it.
+      routes[`/:id/relationships/${dasherizedName}`] = async (request: OrmRequest$, { filter }: { [key: string]: unknown } = {}) => {
         const record = await store.find(model, getId(request.params)) as OrmRecord | undefined;
         if (!record) return 404;
+        if (isDenied(filter, record)) return 404;
 
         const relatedData = record.__relationships[relationshipName];
         const baseUrl = getBaseUrl(request);
@@ -551,32 +1539,140 @@ export default class OrmRequest extends Request {
       };
     }
 
-    // Catch-all for invalid relationship names on related resource route
-    routes[`/:id/:relationship`] = async (request: OrmRequest$) => {
-      const record = await store.find(model, getId(request.params));
-      if (!record) return 404;
+    // Catch-alls for invalid relationship names. Every valid relationship was
+    // registered above, so reaching either of these means the relationship does
+    // not exist and the answer is 404 regardless of the record.
+    //
+    // These deliberately carry NO access check and no store lookup. An earlier
+    // revision of #190 added `if (isDenied(filter, record)) return 404` here for
+    // symmetry with the seven real surfaces, but both branches returned 404, so
+    // the guard was unobservable by construction -- a mutation deleting it
+    // survived the entire suite because no test that could distinguish it can
+    // exist. Unkillable code in an authorization diff reads as coverage and is
+    // not, so it is gone; skipping the lookup also removes the timing difference
+    // between an existing and a missing parent.
+    //
+    // IF THIS ROUTE EVER RETURNS ANYTHING OTHER THAN A CONSTANT 404, it becomes
+    // the eighth surface and must filter the parent first, exactly like
+    // `/:id/{relationship}` above.
+    routes[`/:id/:relationship`] = async () => 404;
+    routes[`/:id/relationships/:relationship`] = async () => 404;
 
-      // If we reach here, relationship doesn't exist (valid ones were registered above)
-      return 404;
+    return routes;
+  }
+
+  auth(request: OrmRequest$, state: { [key: string]: unknown }): number | undefined {
+    // A consumer `access()` that throws is a DENIAL, matching `isDenied` one
+    // layer down. Unguarded it propagates to express's default handler, which
+    // answers 500 -- and the documented sample itself can throw
+    // (`request.originalUrl.split(...)` when originalUrl is absent), so the
+    // failure mode is reachable by following the docs.
+    // -------------------------------------------------------------------------
+    // #202 -- hand the consumer the STRUCTURAL facts, not just the transport.
+    //
+    // Both members are already in hand here. `model` is `this.model`, the name
+    // setup-rest-server mounted this route for; `operation` is the SAME
+    // `methodAccessMap` lookup the permission-array branch at the bottom of
+    // this method performs, so the predicate form and the array form cannot
+    // answer differently about the same request.
+    //
+    // NEITHER IS DERIVED FROM THE REQUEST TARGET, and that is the whole point.
+    // Deriving `model` here from `request.baseUrl` (or from the mounted route
+    // name, or from `getPluralName(this.model)`) would move all five fail-open
+    // variants listed in this file's header OUT of the consumer and INTO the
+    // framework, where every consumer inherits them at once. `this.model` is
+    // assigned once at mount time and no request can influence it.
+    //
+    // `operation` is left UNDEFINED for a method with no entry in
+    // `methodAccessMap`, rather than defaulted. Express delivers HEAD to the
+    // GET handler, so an unmapped method really does reach this line; a
+    // `?? 'read'` here would hand the consumer a fabricated authorisation fact
+    // and turn an unclassified request into an authorised one. Undefined is
+    // the honest answer.
+    //
+    // `record` is deliberately absent -- see `AccessContext` in
+    // src/types/orm-types.ts. Nothing is fetched at this point and adding a
+    // lookup here would put a store read in the middle of an authorization
+    // path. The function return shape below IS the per-record hook.
+    //
+    // -------------------------------------------------------------------------
+    // #236 -- `recordId`, the DECODED route-parameter id, for the same reason.
+    //
+    // WHICH RECORD is the third structural fact the framework already holds and
+    // the consumer was left to re-derive, and re-deriving it failed OPEN. The
+    // documented sample compared `request.path` -- the RAW, undecoded pathname
+    // -- against a literal `/archived`, while the router DECODES `:id`. So
+    // `GET /owners/%61rchived` walked past the deny and was dispatched as the
+    // record `archived`: 200 with the record in full, and DELETE answered 204
+    // with the record destroyed, unauthenticated. Four spellings measured, all
+    // four through; 255 non-canonical spellings of that 8-character id decode
+    // to the same key, so this was never a deny-list of one.
+    //
+    // TWO CONSUMER-SIDE NORMALISATIONS WERE MEASURED WRONG IN OPPOSITE
+    // DIRECTIONS, which is the argument for doing it once, here.
+    // `.toLowerCase()` case-folds a route-parameter VALUE on the axis that
+    // governs literal SEGMENTS: with a distinct owner seeded at `ARCHIVED`,
+    // `GET /owners/ARCHIVED` was a false DENY on the wrong record and
+    // `GET /owners/%41RCHIVED` a false ALLOW on that same one.
+    // `decodeURIComponent(request.path)` decodes THEN splits while the router
+    // splits THEN decodes, so it over-denied `/owners/archived%2fx` -- 403 for
+    // a genuinely distinct record. Failing closed there was luck, not design.
+    //
+    // `getId(request.params)` AND NOT `request.params.id`, for exactly the
+    // reason `operation` is a `methodAccessMap` lookup: it is the SAME single
+    // coercion the store lookup one layer down performs, so the predicate and
+    // the dispatch cannot disagree about which record a request addresses.
+    // The raw string would reintroduce that divergence on hex-shaped ids --
+    // `GET /animals/0x2391` looks up record `9105`.
+    //
+    // NOTHING HERE PARSES THE REQUEST TARGET EITHER. `request.params` is what
+    // the router matched, so a mount prefix, an absolute-form target, a query
+    // string or a case-varied mount cannot move this value -- the same
+    // guarantee `model` carries, by the same means.
+    //
+    // `null` and not `undefined` on a collection route, so the KEY IS ALWAYS
+    // PRESENT -- the rule `operation`'s own docblock already establishes. A
+    // context reaching a predicate WITHOUT the key therefore did not come from
+    // here; it was hand-assembled by a caller resolving the predicate through
+    // `Orm.instance.getAccess()`, and that absence stays deniable only because
+    // `auth()` never produces it.
+    // -------------------------------------------------------------------------
+    const context: AccessContext = {
+      model: this.model,
+      operation: methodAccessMap[request.method],
+      recordId: request.params && 'id' in request.params ? getId(request.params) : null,
     };
 
-    // Catch-all for invalid relationship names on relationship linkage route
-    routes[`/:id/relationships/:relationship`] = async (request: OrmRequest$) => {
-      const record = await store.find(model, getId(request.params));
-      if (!record) return 404;
+    let access: AccessMethod;
+    try {
+      access = this.access(request, context);
+    } catch (error) {
+      // Same reasoning as `isDenied`: fail closed, but say so. An `access()`
+      // that throws denies EVERY request to the collection, and a silent 403
+      // wall is the hardest possible thing to diagnose from the outside.
+      log.error?.(`[@stonyx/orm] access() threw for model "${this.model}" -- denying. ${error instanceof Error ? error.message : String(error)}`);
 
-      return 404;
-    };
+      return 403; // Forbidden
+    }
 
-    return routes;
-  }
+    // THE READING OF THE RETURN SHAPE LIVES IN ONE PLACE (#234).
+    //
+    // It used to be inline here, and it was the only copy, which was fine while
+    // `auth()` was the only thing that had to ask. It is not any more: the
+    // linkage path has to ask model X's predicate about model X's records while
+    // servicing a request routed to model Y, and a second inline copy of these
+    // six branches would be a second authorization vocabulary -- one that can
+    // drift, and that reviewers would have to notice had drifted. The branch
+    // order in `interpretAccess` is this block, moved, not rewritten.
+    const verdict = interpretAccess(access, methodAccessMap[request.method]);
 
-  auth(request: OrmRequest$, state: { [key: string]: unknown }): number | undefined {
-    const access = this.access(request);
+    if (!verdict.granted) return 403;
+
+    // The function return shape is the per-record hook, and `state` is the
+    // whole transport for it: @stonyx/rest-server memoises one state object per
+    // request and hands the same one to `auth()` and to the handler.
+    if (verdict.filter) state.filter = verdict.filter;
 
-    if (!access) return 403;
-    if (Array.isArray(access) && !access.includes(methodAccessMap[request.method])) return 403;
-    if (typeof access === 'function') state.filter = access;
     return undefined;
   }
 }