@@ -28,6 +28,7 @@ jobs:
       - run: bun install --frozen-lockfile
       - run: bun run typecheck
       - run: bun run test
+      - run: bun run test:e2e
   node:
     runs-on: ubuntu-latest
     strategy:
 
@@ -5,217 +5,48 @@ All notable changes to `@corbits/memory` are documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
-## [Unreleased]
+## [0.2.0] — 2026-09-25
 
 ### Changed
 
-- `@intx/*`, `drizzle-orm`, `hono`, `hono-openapi` and `postgres` are peer
-  dependencies; the host supplies them. `engines` is removed.
+- **Breaking:** `@intx/*`, `drizzle-orm`, `hono`, `hono-openapi` and
+  `postgres` are peer dependencies; the host supplies them. `engines` is
+  removed.
+- **Breaking:** `createMemoryRoutes({ memory, requireGrant, callerResolver? })`
+  returns the memory routes as a `Hono<TenantEnv>` sub-app with paths relative
+  to its mount point; hosts mount it at `/api/tenants/:tenantId/memory`. It
+  replaces `createMemory({ app, callerResolver })` and `registerMemoryRoutes`,
+  and `RouteDeps` no longer carries `grants`. `createMemory` only builds the
+  plane.
+- **Breaking:** the package root exports only the public API. Internal
+  services and helpers (transform, retention, feed, share materialization,
+  corroboration, embed model registry, degrade metrics, FTS helpers), the test
+  fakes, and `resolveGrantConfig` are no longer exported. The distiller stays
+  at `@corbits/memory/distiller` and migrations at
+  `@corbits/memory/migrations`.
+- **Breaking:** `runMemoryMigrations(config, { schema, ftsLanguage })` takes
+  the same `DBConfig` as Interchange `runMigrations` instead of a database
+  URL. `schema` names the host schema holding Interchange's `tenant` and
+  `principal` tables (the value passed to `runMigrations`, e.g. `"public"`);
+  memory's tables stay in the `memory` schema. `ftsLanguage` is required, and
+  the runner no longer reads `FTS_LANGUAGE` from the environment or accepts a
+  `log` option.
+- Every migration file is idempotent and replayed on each run, with a 5 s
+  `lock_timeout`. The `memory._migrations` ledger is dropped.
 - `MEMORY_GRANT_REQUIREMENTS` is read from `package.json`
   `interchange.grantRequirements`, now the only declaration.
   `MEMORY_CAPABILITY_IDS` is typed `string[]`.
-- `createMemoryRoutes({ memory, requireGrant, callerResolver? })` returns the
-  memory routes as a `Hono<TenantEnv>` sub-app with paths relative to its
-  mount point; hosts mount it at `/api/tenants/:tenantId/memory`. It replaces
-  `createMemory({ app })` and `registerMemoryRoutes`, and `RouteDeps` no
-  longer carries `grants`. `createMemory` only builds the plane.
-- The package root exports only the public API. Internal services and
-  helpers (transform, retention, feed, share materialization, corroboration,
-  embed model registry, degrade metrics, FTS helpers), the test fakes, and
-  `resolveGrantConfig` are no longer exported. The distiller stays at
-  `@corbits/memory/distiller` and migrations at `@corbits/memory/migrations`.
-- `runMemoryMigrations(config, { schema, ftsLanguage })` takes the same
-  `DBConfig` as Interchange `runMigrations` instead of a database URL.
-  `schema` names the host schema holding Interchange's `tenant` and
-  `principal` tables (the value passed to `runMigrations`, e.g. `"public"`);
-  memory's tables stay in the `memory` schema. `ftsLanguage` is required, and
-  the runner no longer reads `FTS_LANGUAGE` from the environment or accepts a `log` option.
-  Every migration file is idempotent and replayed on each run. The
-  `memory._migrations` ledger is dropped.
-
-### Added
-
-- Retention HTTP routes (CL-6288): `POST …/memory/documents/:documentId/forget`
-  (tombstone, grant `memory:forget`), `POST …/memory/documents/:documentId/purge`
-  (hard delete, grant `memory:purge`), and
-  `POST …/memory/versions/:versionId/retention-class` (grant `memory:forget`).
-  Forget and purge are separate routes with separate grant actions — never one
-  route with a boolean flag — and both are refused with 403 unless the caller
-  is the document/version's creator, independent of any share grant that lets
-  them merely see it. `sweepEphemeral` stays off the HTTP surface (maintenance
-  sweep, not a user action); a host schedules it on its own cron against the
-  in-process `Memory`. New `memory:forget` / `memory:purge` grant requirements
-  (`source: "creator"`) and `capabilityIdsForSurface()` so distiller/tools
-  installs no longer pick up routes-only capabilities by accident.
-- `RouteDeps.callerResolver` — an
-  optional host-supplied resolver from a request to a `{ tenantId,
-  principalId }` scope, for a caller that never goes through the host's
-  tenant-session middleware (e.g. a workflow-run child authenticating with
-  its own sidecar bearer token). Unset by default: every route still reads
-  identity from `c.get("principal")` exactly as before. When set, the
-  resolved identity is seated as the request's principal/tenant ahead of
-  `grantGuard`, so the same `requireGrant` authorization path applies to a
-  machine caller — never a separate, weaker one. Identity from the resolver
-  always wins over anything a request body claims. The resolver's return
-  value is parsed with arktype (non-empty `tenantId`/`principalId`); a
-  resolver returning a malformed identity is rejected with `500` (a host
-  bug), never seated as a garbage scope.
-
-### Fixed
-
-- `loadMemoryConfig` / `EngineConfig.embed` no longer requires an embed
-  endpoint: a host with a pgvector Postgres and no embed endpoint now
-  constructs and serves `add` + lexical `search`. Dense retrieval is skipped
-  (not attempted-and-failed) and `search` reports
-  `degraded: ["dense_unavailable", "lexical_only"]` so the state stays
-  observable (CL-6287). **Migration note:** `dense_unavailable` is now also
-  emitted on every search for a deliberately-unconfigured engine, not only on
-  a transient dense-retrieval failure — a host with an existing alert rule
-  keyed on `dense_unavailable` alone should also check for `lexical_only` in
-  the same `degraded` array to distinguish "opted into lexical-only" from an
-  actual regression.
-- `add`'s `degraded` is now a reason array (`["embed_unavailable"]` and/or
-  `["embed_unavailable", "lexical_only"]`), matching `search`'s shape —
-  previously a bare boolean, which made it impossible to write one
-  "is this response degraded" check across both verbs (CL-6287). **Breaking
-  if a host coded against the boolean:** `degraded: true` is now
-  `degraded: [...]`; check array presence/length instead of truthiness (both
-  are still falsy/omitted when the document captured cleanly).
-- `Memory.capabilities.embeddingsConfigured` (and the underlying
-  `DocumentStore.capabilities`) let a host learn recall is lexical-only at
-  construction time, without issuing a search first (CL-6287).
-- Feed `nextCursor` advances past the examined raw page after grant-tag
-  post-filter (a fully denied page no longer stalls the consumer forever).
-- Distiller default system prompt uses the configured `agentId` for
-  `exclude_generator` / `generator_agent_id` (not only the package default id).
-
-### Changed
-
-- **Breaking:** `createResidentDistiller` runs on a `mail` trigger (CL-8798).
-  `CreateResidentDistillerOpts.cron` is replaced by a required `mailTo`, the
-  address the workflow's run is triggered at, and
-  `RESIDENT_DISTILLER_CRON_DEFAULT` is removed. The `schedule` trigger was
-  reserved with nothing behind it in `@intx/workflow`, so the distiller never
-  ticked, and 0.4.0 rejects it at `defineWorkflow`. Ticking is the host's job:
-  mail `mailTo` on a schedule, e.g. from `@corbits/cron`.
-- `@intx/*` dependencies move to 0.4.0 (CL-8797).
-- **Product narrative:** default path is **add → ingest elements → process**
-  (one host pipeline). Pull feed + `createResidentDistiller` are optional
-  multi-writer / backfill process helpers, not the primary ingest story.
-  See `PRODUCT.md`, `docs/DISTILLER.md`, `docs/FEED.md`.
-- **Breaking:** Postgres schema renamed from `knowledge` to **`memory`**. Fresh
-  installs only — drop/recreate the old schema (or rename) on existing DBs.
-  Citation `open.type` is now `"memory"`.
-- **Breaking:** env var is `DATABASE_URL` (was `KNOWLEDGE_DATABASE_URL`); pass
-  `memory.databaseUrl` on config as an alternative. No deprecated alias.
-- `add` (plane + HTTP) returns `{ documentId, versionId }` so share-grant audit
-  and provenance can name the version that carried the write.
-- Interchange `defineTool` factories at `@corbits/memory/tools` (`memory_add`,
-  `memory_search`, `memory_list`) — HTTP clients for mounted hub routes with
-  install env `memoryBaseUrl` / `memoryTenantId` / `memoryAuthToken`. Declared
-  via `package.json` `interchange.tools` and `exports["./tools"]`.
-
-### Previously
-
-- **Breaking:** package and public surface renamed from `@corbits/knowledge-engine`
-  to `@corbits/memory`. Public APIs: `createMemory` (optional `app` registers HTTP),
-  `registerMemoryRoutes`.
-
-  `createMemory`, `loadMemoryConfig`, `runMemoryMigrations`, `Memory`,
-  `MemoryConfig`, `MemoryError`. HTTP paths are under
-  `/api/tenants/:tenantId/memory/`; grants are
-  `memory:add` / `memory:search`; access tags use `memory.owner:` / `memory.tenant:`
-  / `memory.space:`. Postgres schema is `memory`.
-- **Breaking:** config reads `DATABASE_URL` (was `KNOWLEDGE_DATABASE_URL`) for
-  the engine's own pgvector Postgres connection.
-- **Breaking:** memory plane surface is `add` / `search` / `list` with
-  `principalId` + `tenantId` only. Inference is host-owned (no answer endpoint
-  or personal-memory side-channel on the plane).
-
-- **Breaking:** HTTP routes are `POST /api/tenants/:tenantId/memory/add`,
-  `POST /api/tenants/:tenantId/memory/search`,
-  `GET /api/tenants/:tenantId/memory/list` (inherits hub `resolveTenant`).
-  Old unscoped `/api/memory/*` paths are not mounted.
-- **Breaking:** grant actions are `add` and `search` (was `capture` / `find` /
-  knowledge `search`). `list` uses the `search` grant. Capability resource is
-  `memory`. Document-tag checks use action `search`.
-- **Breaking:** `add` returns `{ documentId }`; search body uses `limit` (not `k`);
-  search wire uses `items` (not `hits`).
-- **Breaking:** document access is Interchange **grant tags** (`accessTags` +
-  creator-always + host `GrantStore`), not the visibility-mode / block-list
-  mini-ACL. Share sugar only mints tags. See `docs/AUTHZ-DOCUMENT-ACCESS.md`.
-- **Breaking:** Postgres baseline is two files (`0001_extensions` +
-  `0002_memory_baseline`) with `access_tags` and no `visibility_*` columns.
-  Fresh installs only — drop/recreate the `memory` schema on existing DBs.
-- **Breaking:** `grantStore` + `conditionRegistry` are top-level `createMemory`
-  options (no nested `grants: { … }`).
-
-### Added
-
-- Optional `TextExtractor` + `file` XOR `content` on `add`
-- `share` sugar on `add` (maps to access tags; principals also materialize
-  grants when the host grant store is writable); `grantsMaterialized` on the
-  add result when peers were requested
-- `access_tags` on `memory.document` (baseline schema)
-- Claim-bearing schema, temporal model, transform/replay, share grants
-  (resident memory distillation foundation — CL-5865/5866/5872/5873)
-- Living relevancy: corroboration factor from supports/contradicts edges;
-  strong evidence gate (CL-5867). See `docs/RELEVANCY.md`
-- Capture feed: `memory.feed` + `GET .../memory/feed` with `feed_seq` cursor
-  (CL-5868). See `docs/FEED.md`
-- **Wire attribution (CL-5870):** search hits carry additive `attribution`
-  (versionId, provenance, source/temporal class, createdByKind,
-  generatorAgentId, occurredAt/validUntil, corroboration counts, derivedFrom)
-- **Retention (CL-5871):** `retention_class` on versions; plane APIs
-  `deprecateVersion` / `tombstoneDocument` / `hardDeleteDocument` /
-  `sweepEphemeral` / `setRetentionClass`; `includeDeprecated` on search.
-  See `docs/RETENTION.md`
-- **Resident distiller (CL-5869):** `@corbits/memory/distiller` —
-  `createResidentDistiller({ inference })` schedule workflow + `runDistillTick`
-  + `buildDistilledClaim`. Claim-aware `add` (generator_agent_id, provenance,
-  derived_from, …). `memory_feed` tool. Feed entries include `accessTags`.
-  See `docs/DISTILLER.md`
-
-### Removed
-
-- Host-injected generate path on the plane (use host inference + `add` / `search`)
-
-## [0.1.2] — 2026-07-31
-
-### Added
-- Public `createMemory` export for out-of-band capture and search (CLI seeders, batch ingesters, tests) without mounting HTTP routes (`#8`)
-
-### Fixed
-
-- Fail-closed document ACL: withhold hits whose `acl_block` is unreadable or non-string (`#9`)
-- Timeline titles use the same document ACL post-filter as search (`#12`)
-- Return 401 (not 500) when the host has not resolved a principal (`#7`)
-- ACL wiring tests mock `sql.unsafe` for FTS verification so the suite stays green under the search path
-
-## [0.1.1] — 2026-07-31
-
-### Added
-
-- Halfvec expression indexes for embedding models above 2000 dimensions (up to 4000); dense search ORDER BY uses the same expression as the index (`#1`)
-- Configurable lexical FTS language (`FTS_LANGUAGE` / `EngineConfig.ftsLanguage`) with parse, catalog verify, and rebuild recipe; public `DEFAULT_FTS_LANGUAGE` export (`#4`)
-- Named chunk FK with `ON DELETE CASCADE` and composite tenant index on per-model embedding tables created at activation (`#3`)
-
-### Changed
-
-- Dense search scales `hnsw.ef_search` to the overfetch limit (floor 40) and probes `hnsw.iterative_scan = relaxed_order` once per process when available (`#2`)
-- Package install docs point at `@corbits/memory` (`#6`)
-
-### Fixed
 
-- Rerank document truncation to fit cross-encoder token limits (`#10`)
-- Degrade metrics: surface full rerank failure detail and count DegradeFlags so silent degradation is visible (`#11`)
+### Upgrading from 0.1.0
 
-## [0.1.0] — 2026-07
+- Call the new `runMemoryMigrations` once; it upgrades a 0.1.0 database in
+  place with no data loss.
+- After the upgrade, do not run 0.1.0 against the database: its ledger is
+  gone, so there is no downgrade.
 
-### Added
+## [0.1.0] — 2026-09-25
 
-- Initial cut: mountable memory capture + search SDK for Interchange hubs
+First release on npm.
 
-[0.1.2]: https://github.com/corbitsdev/corbits-memory/compare/v0.1.1...v0.1.2
-[0.1.1]: https://github.com/corbitsdev/corbits-memory/compare/v0.1.0...v0.1.1
-[0.1.0]: https://github.com/corbitsdev/corbits-memory/releases/tag/v0.1.0
+[0.2.0]: https://github.com/corbitsdev/corbits-memory/releases/tag/v0.2.0
+[0.1.0]: https://www.npmjs.com/package/@corbits/memory/v/0.1.0
@@ -22,11 +22,11 @@ Postgres. See `IMPLEMENTATION.md` for how the pieces fit together.
 ## Running the tests
 
 ```bash
-bun run typecheck && bun run test
+bun run typecheck && bun run test && bun run test:e2e
 ```
 
-- `bun run test` runs the unit suite in `src/` and the end-to-end suite in
-  `tests/`. The end-to-end tests drive the mounted routes and migrations
+- `bun run test` runs the unit suite in `src/`; `bun run test:e2e` runs the
+  end-to-end suite in `e2e/`. The end-to-end tests drive the mounted routes and migrations
   against a real pgvector Postgres: set `TEST_DATABASE_URL` to a server the
   tests can create and drop databases on (for `docker compose up -d`,
   `postgres://memory:memory-dev-password@localhost:5434/memory`). Each suite
@@ -53,18 +53,16 @@ numbered file, because a guarded `ADD CONSTRAINT` keeps the old definition.
   changed; link any relevant issue.
 - Make sure `bun run typecheck && bun run test` pass before requesting review.
 
-## Commit messages
-
-- Present tense, imperative mood: "Add X", "Fix Y", "Harden Z" — not "Added"
-  or "Fixes".
-- No issue-tracker ticket references (e.g. `CL-1234`) in commit messages,
-  code, or comments — commit messages should be self-explanatory without an
-  external ticket.
-- One logical change per commit where practical; a commit message describes
-  the change, not the task that produced it.
-
 ## Contributor License Agreement
 
 Contributions require agreeing to the project's CLA — see `CLA.md`. The CLA
 bot will comment on your first PR with instructions if you haven't signed
 yet.
+
+## Commit messages
+
+Commit subjects and PR titles follow [Conventional Commits](https://www.conventionalcommits.org): `feat`, `fix`, `refactor`, `test`, `docs`, `build`, `ci`, `perf`, and `chore(release): x.y.z` for releases.
+Add `!` only for public API breaks: removed or renamed exports, changed signatures, newly required params. Peer and dependency range changes are `build(deps):` with no `!`.
+Keep subjects imperative, lowercase after the colon, 72 characters or less, and free of ticket IDs.
+No ticket IDs in code or comments either. One logical change per commit where practical; describe the change, not the task that produced it.
+Every PR links its issue with a `Closes <issue id>` line in the PR body.
@@ -64,7 +64,7 @@ creator (`created_by_principal_id` — the document's first version for
 independent of any share grant. A peer who can search a shared document gets
 403 on `forget`/`purge`/`retention-class` for it. See
 `src/services/retention-ownership.ts` and the ownership tests in
-`tests/grants.test.ts` / `tests/caller-resolver.test.ts`.
+`e2e/grants.test.ts` / `e2e/caller-resolver.test.ts`.
 
 **`sweepEphemeral` stays off the HTTP surface.** It is a maintenance sweep —
 "deprecate every ephemeral version past its TTL for this tenant" — not
 
@@ -12,7 +12,7 @@ import {
   seedPrincipal,
   testDatabaseUrl,
   type TestDb,
-} from "./lib/db-harness.ts";
+} from "./helpers.ts";
 
 describe.skipIf(testDatabaseUrl() === undefined)("add and search", () => {
   let db: TestDb;
 
@@ -15,7 +15,7 @@ import {
   seedPrincipal,
   testDatabaseUrl,
   type TestDb,
-} from "./lib/db-harness.ts";
+} from "./helpers.ts";
 
 // A machine caller (e.g. a workflow run) never passes the host's session
 // middleware; the host's callerResolver names its tenant and principal.
 
@@ -12,7 +12,7 @@ import {
   seedPrincipal,
   testDatabaseUrl,
   type TestDb,
-} from "./lib/db-harness.ts";
+} from "./helpers.ts";
 
 describe.skipIf(testDatabaseUrl() === undefined)("synthesized context rows", () => {
   let db: TestDb;
 
@@ -12,7 +12,7 @@ import {
   seedPrincipal,
   testDatabaseUrl,
   type TestDb,
-} from "./lib/db-harness.ts";
+} from "./helpers.ts";
 
 describe.skipIf(testDatabaseUrl() === undefined)("distill tick", () => {
   let db: TestDb;
 
@@ -6,7 +6,7 @@ import {
   embedTexts,
   probeEmbedDims,
 } from "../src/core/embed-client.ts";
-import { startHttpStub } from "./lib/http-stub.ts";
+import { startHttpStub } from "./helpers.ts";
 
 const stub = startHttpStub();
 afterAll(() => stub.stop());