US09-03: Write the Architecture Overview

.dockerignore

@@ -9,6 +9,7 @@
 !photo_pipeline
 !migrations
 !frontend
+!docs
 !docker
 
 # Nothing generated, even under an allowed directory.

.gitattributes

@@ -0,0 +1,9 @@
+# Vendored third-party browser bundles (US09-01). Minified upstream files carry
+# trailing whitespace and very long lines; they are not ours to reformat, and the
+# submit gate's `git diff --check` would refuse them forever. Their integrity is
+# controlled where it belongs instead: a pinned version and a recorded sha256 in
+# frontend/js/vendor/VERSIONS.json, asserted by tests/integration/test_documentation.py.
+#
+# Only the whitespace check is turned off. They stay text, so a credential scan or a
+# search still reads them.
+frontend/js/vendor/*.js -whitespace linguist-vendored

.gitea/workflows/test.yml

@@ -49,3 +49,55 @@ jobs:
 
       - name: Application suite
         run: work_item/scripts/python -m pytest tests -q
+
+  # The deployed container, verified the way the host application is (US08-05). It
+  # runs on `main` only — a pull request has nothing published to upgrade *from*, and
+  # building two images per push would pay for that on every commit. Deploy waits for
+  # this whole workflow, so a red container gate is a deploy that does not happen.
+  container:
+    if: gitea.event_name == 'push'
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        # The upgrade journey builds the previous commit's tree when no published
+        # image is named, so the history has to be there.
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 2
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.12'
+
+      - name: Install the application and its test dependencies
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install -e '.[test]'
+          python -m playwright install --with-deps chromium
+
+      - name: Container acceptance gate
+        # One command: it builds the image, provisions the composition against a
+        # temporary fixture library and an isolated volume, runs the phase_h journeys,
+        # destroys the stack, and writes the evidence. Any skipped check fails it.
+        env:
+          # The fixture libraries are bind-mounted into the containers, so this path
+          # has to be one the Docker daemon can see. On a runner that talks to a
+          # sibling daemon, point it at a shared host path instead of the workspace —
+          # an unshared path arrives as an empty mount and the journeys fail on the
+          # scan, which is the symptom to recognise.
+          PHOTO_PIPELINE_DATA_DIR: ${{ gitea.workspace }}/gate-data
+          PHOTO_PIPELINE_LIBRARY_ROOTS: ${{ gitea.workspace }}/gate-library
+          PHOTO_PIPELINE_TEST_MOUNT_BASE: ${{ gitea.workspace }}/gate-mounts
+        run: |
+          mkdir -p "$PHOTO_PIPELINE_LIBRARY_ROOTS" "$PHOTO_PIPELINE_TEST_MOUNT_BASE"
+          python -m photo_pipeline container-gate --output gate-evidence
+
+      - name: Keep the evidence
+        # Retained per run, and retained on failure especially: the logs are the only
+        # account of what the containers did.
+        if: always()
+        uses: actions/upload-artifact@v3
+        with:
+          name: container-gate-${{ gitea.sha }}
+          path: gate-evidence

Dockerfile

@@ -70,6 +70,8 @@ COPY pyproject.toml alembic.ini README.md ./
 COPY photo_pipeline ./photo_pipeline
 COPY migrations ./migrations
 COPY frontend ./frontend
+# The manuals are served by the application itself (US09-01), so they ship with it.
+COPY docs ./docs
 COPY docker/entrypoint.sh docker/healthcheck.sh /usr/local/bin/
 
 # Runtime dependencies only: the `test` extra (pytest, playwright) and the `vision`

README.md

@@ -241,6 +241,40 @@ image. The stack is managed by Portainer from git (`docker-compose.yml`), and th
 values live in the Portainer stack's environment, so rotation is one place and a
 repository read discloses nothing. A test asserts the workflows never name them.
 
+## Container acceptance gate (US08-05)
+
+The deployed container is verified the way the host application is, by one command:
+
+```bash
+work_item/scripts/python -m photo_pipeline container-gate --output gate-evidence
+```
+
+It builds the image, provisions the composition against a **temporary fixture library
+on a bind mount and an isolated data volume**, runs the `phase_h` journeys against it,
+destroys every stack afterwards, and writes `release-report.json`, `logs/container.log`,
+and `CHECKSUMS.sha256` into the evidence directory. It exits non-zero when a journey
+fails, when the story matrix has a hole, **or when any check skipped at all** — unlike
+the release gate, this one accepts no environment excuse: a run that never reached the
+containers proved nothing about them.
+
+The journeys (`tests/e2e/test_phase_h_container.py`) are:
+
+| journey | what it proves |
+|---|---|
+| browser | discovery, duplicate review, analysis, album proposal, rename, upload preflight, and archive views, driven through the containerized frontend — including a rename that really moves the operator's folder on the bind mount |
+| upgrade | the previous version's image runs first, then this one against the same volume: schema at the new head, assets, analysis results, job progress, the unapplied rename plan, and the thumbnail cache all survive |
+| restart | `docker kill` on both containers mid-job; the job resumes, every photo ends with exactly one stored result, and only the in-flight item ever reaches the provider twice |
+| security | no session refused, a forged `X-Forwarded-Host` cannot smuggle an allowed hostname past the check, a symlink out of the mounted library is refused, an unmounted library root refuses startup, and no secret appears in `docker compose logs` |
+
+The upgrade journey builds the previous commit's tree when no published image is named;
+point it at the real one with `PHOTO_PIPELINE_PREVIOUS_IMAGE`. On a Docker VM (Colima,
+Docker Desktop) the fixture library must live on a shared path — it defaults to
+`~/.cache/photo-pipeline`, overridable with `PHOTO_PIPELINE_TEST_MOUNT_BASE`.
+
+CI runs this gate as the `container` job of **Test** on pushes to `main` and keeps its
+evidence as a run artefact. Deploy waits for the whole `Test` workflow, so a red
+container gate is a publish that does not happen.
+
 ## Testing
 
 One offline command runs the whole suite (unit, integration, and browser

delivery_backlog/E09-documentation.md

@@ -0,0 +1,77 @@
+# E09 — Product Documentation
+
+Concept phase: none. Like [E08](E08-container-deployment.md), this epic is a delivery
+addition rather than a product-scope change: the same application, documented well
+enough that somebody who did not build it can install it, understand it, and operate it
+without reading the source.
+
+It does not change the product scope in
+[`INTEGRATED_PIPELINE_CONCEPT.md`](../INTEGRATED_PIPELINE_CONCEPT.md). No safety
+invariant moves, no schema changes, no new runtime capability. The one change to the
+running application is a documentation view and the static assets it needs.
+
+## Why the application serves its own documentation
+
+The manuals describe an application that is reached over HTTP behind an access secret.
+Documentation that lives only in the repository is unreachable from the deployment it
+describes: an operator who has just been handed a URL and a secret has no Gitea account
+in front of them. So the same markdown files are both the repository's documentation and
+the deployment's `/docs` view, and neither is a copy of the other.
+
+## Decisions made in this epic
+
+**Markdown is the source.** Everything is written as markdown under `docs/`, so it is
+reviewable in a diff, readable on Gitea, and renderable in the app. No documentation
+format that only a tool can read.
+
+**The renderer is vendored, not written and not fetched.** `marked` (MIT, no
+dependencies, ships an ES module) is pinned and committed under
+`frontend/js/vendor/`. A CDN is not an option: the application is deployed to a network
+whose outbound access is not assumed, and `default-src 'self'` forbids it.
+
+**Diagrams are mermaid, rendered client-side, with the script boundary intact.** The
+claim that mermaid requires `'unsafe-eval'` was tested rather than believed: mermaid 11's
+bundle contains no `eval(` and no `new Function` — only a lodash `Function("return this")`
+global-detection fallback that short-circuits on `globalThis` and never executes.
+Rendered under this application's exact CSP, a flowchart produced a 15.8 KB SVG and
+raised **no `script-src` violation**. What it does raise is `style-src`: mermaid styles
+its output with an injected `<style>` element and `style=` attributes.
+
+So `script-src 'self'` stays exactly as US07-02 and US08-01 left it, and `style-src`
+gains `'unsafe-inline'`. That relaxation is bounded on purpose: with `script-src` intact
+no injected markup can execute, and with `img-src`, `connect-src`, and `font-src` all
+still `'self'`, the CSS-based exfiltration channels stay closed. What remains is
+defacement of a page the operator is already authenticated on.
+
+The rejected alternative is recorded because it may become the better trade later:
+pre-rendering each diagram to a committed SVG with the already-installed Playwright
+(no Node toolchain needed) and serving that, which would leave CSP untouched entirely at
+the cost of a generator and a freshness check. If `style-src 'unsafe-inline'` is ever
+judged too much, that is the migration, and it does not change a single markdown file.
+
+**Screenshots are produced, not pasted.** Every screenshot in the user manual is captured
+by Playwright from the real application against a temporary fixture library, by the same
+kind of code that already drives the browser suites. A screenshot nobody can regenerate
+is a screenshot that silently stops being true.
+
+**The documentation is held to the code by tests.** Configuration keys, CLI commands,
+exit codes, API error codes, job and journal states, and module names all appear in both
+the code and the manuals. Each of those is cross-checked, so the failure mode of stale
+documentation is a red test rather than a misled operator.
+
+## Stories
+
+1. [US09-01 — Serve the documentation inside the application](stories/US09-01-docs-in-app.md)
+2. [US09-02 — Write the installation and operations manual](stories/US09-02-installation-manual.md)
+3. [US09-03 — Write the architecture overview](stories/US09-03-architecture-overview.md)
+4. [US09-04 — Write the user manual with generated screenshots](stories/US09-04-user-manual.md)
+5. [US09-05 — Automate documentation acceptance](stories/US09-05-docs-gate.md)
+
+## Epic outcome
+
+A deployed instance serves its own installation manual, architecture overview, and
+illustrated user manual at `/app/#/docs`, rendered from the same markdown files that are
+readable in the repository. Screenshots are regenerated from the running application,
+every documented command, setting, state, and error code is cross-checked against the
+code that implements it, and one gate fails the build when documentation and application
+disagree.

delivery_backlog/README.md

@@ -2,13 +2,14 @@
 
 This backlog decomposes the phases in
 [`INTEGRATED_PIPELINE_CONCEPT.md`](../INTEGRATED_PIPELINE_CONCEPT.md) into seven
-epics and small, independently verifiable user stories, plus one delivery-format
-epic (E08) that packages the released application as a deployable container.
+epics and small, independently verifiable user stories, plus two delivery-format
+epics: E08 packages the released application as a deployable container, and E09
+documents it for the people who install, operate, and use it.
 
 ## Numbering and file naming
 
-- Epics: `E01` through `E07`, matching concept Phases A through G; `E08` has no
-  concept phase and must not change product scope.
+- Epics: `E01` through `E07`, matching concept Phases A through G; `E08` and `E09`
+  have no concept phase and must not change product scope.
 - Stories: `US<epic>-<sequence>`, for example `US03-02`.
 - Epic files: `E01-<slug>.md`.
 - Story files: `stories/US01-01-<slug>.md`.
@@ -39,6 +40,7 @@ epic (E08) that packages the released application as a deployable container.
 6. [E06 — Archive lifecycle](E06-archive-lifecycle.md)
 7. [E07 — Hardening and release](E07-hardening-release.md)
 8. [E08 — Container deployment](E08-container-deployment.md)
+9. [E09 — Product documentation](E09-documentation.md)
 
 ## Shared definition of done
 

delivery_backlog/stories/US09-01-docs-in-app.md

@@ -0,0 +1,49 @@
+# US09-01 — Serve the Documentation Inside the Application
+
+Epic: [E09](../E09-documentation.md)
+
+As an operator who was handed a URL and an access secret, I want the manuals inside the
+application I am looking at, so that understanding it does not require a repository
+checkout or an internet connection.
+
+## Acceptance criteria
+
+- Documentation lives as markdown under `docs/`, with an index that names every page and
+  the order it is meant to be read in. The same files render on Gitea without
+  modification.
+- The application serves a `/docs` view reachable from the main navigation, listing the
+  pages and rendering the selected one.
+- Rendering uses a vendored, version-pinned markdown library committed under
+  `frontend/js/vendor/`. It is not fetched from a CDN, not bundled by a build step, and
+  not written by hand. The build records the version and a checksum of the vendored file.
+- Diagrams written as ```mermaid``` fenced blocks render as diagrams. Mermaid is vendored
+  the same way.
+- `script-src` in the Content-Security-Policy is unchanged: no `'unsafe-eval'`, no
+  `'unsafe-inline'`. Only `style-src` gains `'unsafe-inline'`, and the reason is recorded
+  where the policy is defined.
+- Links between documents work in the app: a relative `../foo.md#section` link navigates
+  to that page and heading rather than downloading a file or leaving the application.
+  Every heading has a stable anchor, and a deep link to an anchor scrolls to it.
+- An unknown page renders a documentation-specific not-found message with a link back to
+  the index, and never exposes a filesystem path.
+- Documentation is readable without an access secret **or** behind the same session as
+  the rest of the application — whichever is chosen is stated explicitly in the story's
+  implementation notes and covered by a test, because an operator locked out by a
+  configuration mistake is exactly who needs the troubleshooting page.
+- The view works with no outbound network access at all.
+
+## Automated tests
+
+- Unit: heading-anchor slugs (duplicates, punctuation, non-ASCII), and the rewriting of
+  relative markdown links into in-app routes.
+- Integration: every markdown file under `docs/` is reachable from the index; every
+  internal link in every document resolves to a file and, where an anchor is given, to a
+  heading that exists; the vendored library files match their recorded checksums.
+- Browser: the documentation view renders a page, follows an internal link, deep-links to
+  an anchor, renders a mermaid diagram to an SVG, and shows the not-found message for an
+  unknown page — all with **zero console errors and zero CSP violations**, asserted, not
+  eyeballed.
+
+## Dependencies
+
+- None beyond the delivered application.

delivery_backlog/stories/US09-02-installation-manual.md

@@ -0,0 +1,46 @@
+# US09-02 — Write the Installation and Operations Manual
+
+Epic: [E09](../E09-documentation.md)
+
+As somebody installing this application for the first time, I want one manual that takes
+me from nothing to a running instance pointed at my photo library, so that I do not have
+to reconstruct the procedure from the README, the compose file, and the test suite.
+
+## Acceptance criteria
+
+- A **host installation** path: prerequisites and their versions (Python, exiftool,
+  immich-go, Playwright's browser for the test suite), virtual environment, dependency
+  install, `.env`, database migration, starting `serve` and `worker`, and how to verify
+  the install succeeded.
+- A **container installation** path: the published image, the compose file, the data
+  volume, the library bind mount and why it is mounted the way it is, published ports,
+  running behind a reverse proxy with `PHOTO_PIPELINE_ALLOWED_HOSTS` and
+  `PHOTO_PIPELINE_TRUSTED_PROXIES`, and the Portainer/webhook deployment already in use.
+- A **configuration reference**: every `PHOTO_PIPELINE_*` setting with its meaning,
+  default, accepted values, and whether it is a secret. Secrets are described, never
+  exemplified with a real-looking value.
+- A **first run checklist** that ends in a verified state: library discovered, worker
+  claiming jobs, readiness endpoint green, diagnostics clean.
+- **Operations**: upgrading (including what the migration backup does), backup, verify,
+  restore, pruning, diagnostics, and the log locations for each role.
+- **Troubleshooting**: each refusal an operator can hit at startup — every non-zero exit
+  code of the CLI, the trust-boundary refusal, the unmounted library root refusal, the
+  library lock being held, and a legacy CLI still running — with the cause and the fix.
+- The manual states the safety invariants an installer must not work around: one worker
+  writes, `_IGNORE/` is never read, EXIF is verified before upload, and the library lock
+  is authoritative.
+
+## Automated tests
+
+- Every `PHOTO_PIPELINE_*` setting documented exists as a field on `Config`, and every
+  field on `Config` is documented — both directions, so a new setting cannot ship
+  undocumented.
+- Every CLI command and flag named in the manual exists in the argument parser, and every
+  subcommand the parser accepts appears in the manual.
+- Every exit code documented is one the CLI can actually return.
+- No documented example value collides with a real secret pattern (the repository's own
+  credential scanner runs over `docs/`).
+
+## Dependencies
+
+- US09-01

delivery_backlog/stories/US09-03-architecture-overview.md

@@ -0,0 +1,47 @@
+# US09-03 — Write the Architecture Overview
+
+Epic: [E09](../E09-documentation.md)
+
+As a developer or reviewer new to this repository, I want an architecture overview that
+explains what the pieces are and which rules they enforce, so that I can find the right
+module and not violate an invariant I never knew existed.
+
+## Acceptance criteria
+
+- A **context diagram**: the operator, the browser application, the API and worker, the
+  photo library, the Immich server, the vision provider, and the archive destination —
+  with the direction and nature of every interaction, including which ones leave the
+  machine.
+- A **runtime diagram**: the migrate/api/worker composition, the data volume, the library
+  bind mount, the SQLite database, and the process lock — showing what is shared and what
+  is exclusive.
+- A **module map**: every package under `photo_pipeline/` with its responsibility and the
+  boundary it must not cross (which modules may touch the filesystem, which may call an
+  external provider, which own schema).
+- **Key flows**, each as a diagram plus prose: the durable job lifecycle from enqueue to
+  recovery; the rename journal state machine including the rollback and manual-resolution
+  paths; the upload lifecycle through preflight, run, report ingestion, and verification.
+- **The invariants and where they are enforced**, each pointing at the module that owns
+  it: one writer at a time, the library process lock, the path policy and library roots,
+  `_IGNORE/` exclusion, verified EXIF before upload, safety decisions gating the vision
+  provider, and restart-safety of every mutation.
+- **The data model**: the tables, what identity means for an asset, and how a moved file
+  keeps its identity.
+- A short **history and donor** section: what was migrated out of the legacy CLIs, where
+  the archive and its ledger live, and why they are kept.
+- Diagrams are ```mermaid``` blocks so the source is diffable and Gitea renders them
+  natively.
+
+## Automated tests
+
+- Every package and top-level module under `photo_pipeline/` appears in the module map,
+  and every module named in the map exists — a new service cannot appear on the map's
+  blind side.
+- Every job state, journal state, and upload batch state named in the document exists in
+  the code, and every state the code defines is named in the document.
+- Every mermaid block parses (rendered in the browser test without an error node).
+- Every code path referenced by file name exists.
+
+## Dependencies
+
+- US09-01

delivery_backlog/stories/US09-04-user-manual.md

@@ -0,0 +1,52 @@
+# US09-04 — Write the User Manual with Generated Screenshots
+
+Epic: [E09](../E09-documentation.md)
+
+As the person actually sorting a photo library, I want a manual that walks the workflow
+screen by screen and tells me what each refusal means, so that I can use the application
+confidently and know what it is about to do to my files before it does it.
+
+## Acceptance criteria
+
+- One page per workflow stage, in the order the application presents them: discovery and
+  inventory, duplicate review, safety review, analysis, album proposals, renames, upload,
+  archive, and diagnostics. Each page answers the same four questions: what this stage is
+  for, what I have to decide, **what it changes on disk or on the server**, and what it
+  refuses to do.
+- A **guided first pass** that takes a new library from scan to a verified upload, naming
+  the point of no return in each stage and what is reversible after it.
+- Every stage page carries at least one screenshot of the real application showing the
+  state being described.
+- Screenshots are **generated** by a committed Playwright script that seeds a temporary
+  fixture library, drives the application, and writes the images. Regenerating them is one
+  documented command. No screenshot is captured by hand.
+- Screenshots contain no real library paths, no personal photos, and no secrets — the
+  fixture library is synthetic, and this is asserted rather than assumed.
+- An **error and refusal catalogue**: every `error.code` the API can return, with what
+  causes it, what the application did or refused to do, and what the operator should do
+  next. Refusals that protect data (`lock_held`, `rename_recovery_required`,
+  `stale_preflight`, `path_not_allowed`, `host_not_allowed`, `dry_run_not_approved`,
+  conflict codes) are explained as intentional, not as faults.
+- A **recovery** page: an interrupted rename, an uncertain upload, a failed migration, a
+  restored backup — what the application does by itself and what needs a decision.
+- The work-item safety allow list is extended to permit `docs/images/**`, since the
+  repository denies image files by default. This is an explicit, reviewed change, not a
+  quiet one, and it stays narrow enough that a real photo still cannot be committed.
+
+## Automated tests
+
+- Every `error.code` the application can emit is documented, and every code documented
+  exists in the code — both directions.
+- Every image referenced by a documentation page exists, and every image under
+  `docs/images/` is referenced by a page.
+- The screenshot generator runs end to end in the test environment and produces every
+  image the manual references, against a temporary fixture library that is destroyed
+  afterwards.
+- Generated screenshots are checked for library paths outside the fixture root and for
+  the configured secrets.
+- Browser: each stage page renders in the documentation view with its screenshot loaded
+  and no console error.
+
+## Dependencies
+
+- US09-01

delivery_backlog/stories/US09-05-docs-gate.md

@@ -0,0 +1,38 @@
+# US09-05 — Automate Documentation Acceptance
+
+Epic: [E09](../E09-documentation.md)
+
+As a release owner, I want one gate that proves the documentation still describes the
+application, so that a change to the code cannot quietly make the manuals wrong.
+
+## Acceptance criteria
+
+- One documented command runs every documentation check and retains its evidence, in the
+  shape the release and container gates already use.
+- The gate fails when: an internal link or anchor is dead; a page is unreachable from the
+  index; an image is referenced but missing or present but unreferenced; a documented
+  setting, command, exit code, error code, or state does not exist in the code; a code
+  path exists that the documentation is required to cover and does not.
+- The gate regenerates the screenshots and fails when a regenerated image no longer
+  matches the committed one beyond a stated tolerance, so a UI change that invalidates the
+  manual is a red build rather than a discovery months later.
+- The gate renders every documentation page in a real browser and fails on any console
+  error or CSP violation, and asserts that `script-src` contains neither `'unsafe-eval'`
+  nor `'unsafe-inline'`.
+- The checks run on a `phase_i` marker; CI runs the gate, and the earlier epic suites keep
+  running unchanged.
+- The README and the application's documentation index point at each other, so neither is
+  the forgotten copy.
+- Documentation stories are mapped in the story traceability matrix like every other
+  story.
+
+## Automated tests
+
+- The gate's own contract is testable without a browser: a seeded broken link, a missing
+  image, an undocumented error code, and a stale screenshot each fail it, and a clean tree
+  passes.
+- The full documentation suite runs on `phase_i` in CI.
+
+## Dependencies
+
+- US09-01 through US09-04

docs/architecture.md

@@ -0,0 +1,224 @@
+# Architecture
+
+[← Documentation index](index.md)
+
+For whoever has to change this code without breaking somebody's photo library. It
+explains what the pieces are, which rules each one keeps, and where to look when a
+stage refuses.
+
+The product decisions behind all of it live in `INTEGRATED_PIPELINE_CONCEPT.md`; this
+page describes what was built.
+
+## Context
+
+Five things outside the application, and what actually crosses each boundary.
+
+```mermaid
+flowchart LR
+  operator([Operator]):::person -->|browser, one session| app
+  app[Photo Pipeline]:::system -->|read, rename, EXIF write| library[(Photo library<br/>bind mount)]
+  app -->|database, cache, journals, backups| data[(Data directory<br/>local filesystem)]
+  app -->|confirmed-SFW images only| vision[Vision provider]:::ext
+  app -->|immich-go, verified bytes| immich[Immich server]:::ext
+  app -->|copy, verify, then remove| archive[(Archive medium)]
+  classDef person fill:#1f6feb,stroke:#58a6ff,color:#fff
+  classDef system fill:#238636,stroke:#3fb950,color:#fff
+  classDef ext fill:#6e40c9,stroke:#a371f7,color:#fff
+```
+
+| boundary | leaves the machine? | carries |
+|---|---|---|
+| operator → app | no (loopback, or a proxy you configured) | commands against stable ids, never paths |
+| app → library | no | reads, folder renames, EXIF merges |
+| app → data directory | no | SQLite in WAL mode, thumbnails, journals, backups |
+| app → vision provider | **yes** | image bytes of confirmed-SFW canonical assets only |
+| app → Immich | **yes** | the exact verified bytes of an approved album |
+| app → archive medium | no | copies, verified before the source is removed |
+
+## Runtime
+
+Two long-lived processes and a one-shot migration, sharing one database and one
+library.
+
+```mermaid
+flowchart TD
+  browser([Browser]) -->|JSON + SSE| api
+  migrate["migrate<br/>backup, then upgrade"] -->|must exit 0| api
+  migrate --> worker
+  api["api · serve<br/>enqueues, serves, reads"] -->|jobs table| db[(SQLite WAL)]
+  worker["worker<br/>claims and does the work"] -->|jobs table| db
+  worker --> tools["exiftool · vision API · immich-go"]
+  api -.->|api.lock.json| lock{{library lock}}
+  worker -.->|worker.lock.json| lock
+  api --> lib[(library)]
+  worker --> lib
+```
+
+`serve` enqueues and renders; **it does not do the work**. Everything that scans,
+scores, analyses, renames, uploads, or archives happens in the worker, which claims
+a queued job atomically with a fencing token. One mutating job runs at a time, and
+the lock file in the data directory is what makes a second worker impossible rather
+than merely discouraged.
+
+## The modules
+
+| package | owns | must not |
+|---|---|---|
+| `api/` | HTTP surface, error envelope, security policy | contain SQL or business logic |
+| `api/routes/` | one module per resource group, all under `/api/v1` | accept a filesystem path from the browser |
+| `api/security.py` | host/origin/CSRF/session/size policy as one pure `evaluate`, plus the ASGI middleware | be bypassed per-route |
+| `schemas/` | Pydantic request and response contracts | reach the database |
+| `services/` | all domain logic; the only place a decision is made | be imported by the frozen CLI archive |
+| `models/` | SQLAlchemy tables | hold behaviour |
+| `jobs/` | durable job lifecycle, worker loop, lock ranks, handler registry | run work in the API process |
+| `integrations/` | the real outside world: `exiftool`, vision, NSFW model, `immich-go` and its report grammars | be called without a version recorded |
+| `path_policy.py` | the library boundary, `_IGNORE/` exclusion, symlink-escape refusal | be duplicated anywhere |
+| `imaging.py` | the single bounded-decode door | let a caller open an image directly |
+| `faults.py` | the crash barriers the tests fire | read anything but its one env var |
+| `config.py` | typed settings, secrets as `SecretStr` | log or return a value |
+| `db.py` | engine, sessions, WAL and foreign keys, migration entry | be used to bypass a repository |
+
+Services worth knowing by name: `inventory`, `duplicates`, `safety`, `analysis`,
+`albums`/`proposals`/`naming`, `renames`/`rename_apply`/`rename_journal`,
+`uploads`/`upload_batches`/`upload_reports`/`upload_verification`,
+`archives`/`archive_transfer`/`archive_journal`/`restores`/`availability`,
+`thumbnails`, `hashing`, `exif_checkpoint`, `jobs`, `workflow`, `backup`,
+`diagnostics`, `app_lock`, `release`, `benchmarks`, `library`, `legacy_import`.
+
+## The state machines
+
+Three journals decide what a restart is allowed to assume. All three are read from
+the database plus the real world — never guessed from a missing file.
+
+### Durable jobs
+
+```mermaid
+stateDiagram-v2
+  [*] --> queued
+  queued --> running: claimed with a fencing token
+  queued --> cancelled
+  queued --> cancelling
+  running --> succeeded
+  running --> failed
+  running --> cancelling
+  cancelling --> cancelled
+  failed --> retry_queued
+  retry_queued --> running
+  succeeded --> [*]
+  cancelled --> [*]
+```
+
+`succeeded` and `cancelled` are terminal, so a duplicate delivery cannot move a
+finished job. Claiming is compare-and-set on the row version; a worker whose lease
+expired cannot commit after another has taken over.
+
+### The rename journal
+
+The only state machine that moves somebody's folders.
+
+```mermaid
+stateDiagram-v2
+  [*] --> planned
+  planned --> moving
+  moving --> moved
+  moving --> planned: proven untouched
+  moving --> rollback_required
+  moved --> database_updated
+  database_updated --> verified
+  verified --> complete
+  moved --> rollback_required
+  database_updated --> rollback_required
+  verified --> rollback_required
+  planned --> failed
+  failed --> planned
+  rollback_required --> rolled_back
+  complete --> [*]
+  rolled_back --> [*]
+```
+
+Intent is written **before** the disk is touched, which is why `moving` can resolve
+backwards: the evidence decides. `moving`, `moved`, `database_updated`, and
+`rollback_required` are the unsafe states — while any operation sits in one, the
+library may be half-renamed, so unrelated mutations are refused with
+`409 rename_recovery_required` until a person resolves it.
+
+### Upload batches
+
+```mermaid
+stateDiagram-v2
+  [*] --> planned
+  planned --> running
+  running --> succeeded
+  running --> failed
+  running --> cancelling
+  running --> unknown_requires_verification: process died after acceptance
+  cancelling --> cancelled
+  failed --> running: retry
+  cancelled --> running: retry
+```
+
+`unknown_requires_verification` is deliberately **not** restartable. The server may
+already hold the files; the answer is to ask Immich for the recorded SHA-1, not to
+upload again and hope.
+
+Archive transfers use the same shape — `planned → transferring → verified →
+removing → complete` — and the source is removed only after the archived bytes are
+verified.
+
+## Identity and the data model
+
+A path is metadata. The identity is `assets.id`, a UUID that never changes, and
+`asset_paths` records every path an asset has ever had with the reason it changed.
+That is what makes a rename cheap: nothing else in the database has to move.
+
+Tables: `assets`, `asset_paths`, `thumbnails`, `jobs`, `job_items`, `job_events`,
+`safety_reviews`, `analysis_results`, `exif_projections`, `duplicate_clusters`,
+`duplicate_members`, `duplicate_negative_links`, `album_proposals`, `rename_plans`,
+`rename_operations`, `upload_batches`, `upload_items`, `upload_verifications`,
+`archive_locations`, `archive_plans`, `archive_operations`.
+
+Availability is independent of workflow progress: `active`, `archiving`,
+`archived_online`, `archived_offline`, `restoring`, `missing_unexpected`. An
+unmounted archive disk is `archived_offline`, never `missing` — the scanner is not
+allowed to conclude that a photo is gone because a disk is unplugged.
+
+EXIF is a projection with its own verified state (`verified`, `divergent`,
+`failed`): the desired values are written, read back, and compared, and anything the
+stage does not own having changed makes the asset `divergent` and blocks the next
+mutating stage.
+
+## Where each invariant lives
+
+| invariant | enforced in |
+|---|---|
+| `_IGNORE/` is never traversed, counted, or opened | `path_policy.is_excluded`, used by every discovery path |
+| no path outside the library roots is reachable | `path_policy.resolve_in_roots` — it returns the resolved path, because validating one name and opening another is the symlink race |
+| one writer per library | `services/app_lock.py` (an `flock` on a JSON lock file, per role) |
+| one mutating job at a time | `services/jobs.py` lock keys plus `jobs/locks.py` rank ordering |
+| only confirmed-SFW assets reach the vision provider | `services/analysis.py`, re-checked *after* the provider call so a decision that flipped mid-flight discards the result |
+| EXIF is verified, and other fields preserved | `services/exif_checkpoint.py` |
+| uploads carry the exact verified bytes | `services/uploads.py` preflight, re-proved immediately before the uploader runs |
+| an uncertain upload is not a failure | `services/upload_batches.py`, `services/upload_verification.py` |
+| the archive source outlives its copy until verified | `services/archive_transfer.py` |
+| decoding is bounded | `imaging.py` |
+| the trust boundary and CSRF | `api/security.py` |
+
+## Concurrency
+
+Locks are taken broad to narrow — `library → stage/job → album/folder → asset` — and
+never the other way, which is what makes deadlock structural rather than lucky.
+Ranks live in `jobs/locks.py`.
+
+Exactly-once execution is not achievable across SQLite, a filesystem, subprocesses,
+and a remote server. The design promises **at-least-once with idempotent recovery**:
+every handler may run twice, and running twice must not produce two side effects.
+That is why the rename journal records intent before moving, why EXIF writes merge
+and verify, and why upload retries consult both local history and Immich.
+
+## History
+
+This application was extracted from two command-line tools rather than written from
+nothing. Their sources are frozen in `legacy_cli_archive/` with the ledger mapping
+each donated behaviour to the service that now owns it, the characterization tests
+that pinned it, and every intentional difference. Production code must not import
+them; they are provenance and rollback evidence.

docs/index.md

@@ -0,0 +1,35 @@
+# Photo Pipeline documentation
+
+One local application that takes a photo library from discovery to a verified Immich
+upload: duplicate detection, safety review, content analysis, album naming, guarded
+renaming, upload, and archive — one visible, resume-safe workflow.
+
+These pages are readable three ways, and they are the same files each time: in the
+repository under `docs/`, on Gitea, and inside the running application under
+**Docs**. There is no separate copy to fall out of date.
+
+## Read in this order
+
+1. [Overview](overview.md) — what the application does, the stages it moves a photo
+   through, and the rules it will not break.
+2. [Installation and operations](installation.md) — host and container installation,
+   every setting, the first-run checklist, upgrades, backup and restore, and what
+   each refusal at startup means.
+3. [Architecture](architecture.md) — the context and runtime diagrams, what each
+   module owns, the three journals a restart reads, and where every invariant is
+   enforced.
+
+## Being written
+
+The remaining manuals are accepted work, not aspiration; each is a story in
+[E09](https://git.domverse-berlin.eu/domverse/photoanalyzer/src/branch/main/delivery_backlog/E09-documentation.md)
+and will appear here as it lands.
+
+- **User manual** — one page per workflow stage with screenshots of the real
+  application, and a catalogue of every error and refusal (US09-04).
+
+## Conventions
+
+A page tells you what a stage **changes on disk or on the server** before it tells
+you how to run it. Refusals are documented as intentional: this application would
+rather stop and explain than guess about somebody's photographs.

docs/installation.md

@@ -0,0 +1,336 @@
+# Installation and operations
+
+[← Documentation index](index.md)
+
+From nothing to a running instance pointed at your photo library, and everything you
+need afterwards: upgrading, backing up, restoring, and understanding a refusal.
+
+Two ways to install. **Container** is the deployment this project builds and ships;
+**host** is what you want for development or a single machine you already manage.
+Both run the same two processes against the same database.
+
+> Read [what it will not do](overview.md#what-it-will-not-do) first. Several of the
+> steps below only make sense once you know which rules the application is keeping.
+
+## Before you start
+
+| you need | version | why |
+|---|---|---|
+| Python | 3.12 or newer | the application |
+| `exiftool` | 13.x (the image pins `13.25+dfsg-1`) | every EXIF read and write |
+| `immich-go` | 0.32.0 (pinned in the image) | the upload stage only |
+| Docker + the compose plugin | any current | the container installation only |
+| a vision provider key | — | the analysis stage only |
+
+You also need a photo library you can afford to be wrong about. Take a backup of it
+before pointing anything at it for the first time, and consider running with
+[`PHOTO_PIPELINE_REQUIRE_DRY_RUN_APPROVAL`](#the-settings) on.
+
+## Host installation
+
+```bash
+python3.12 -m venv .venv
+.venv/bin/pip install -e ".[vision]"     # drop [vision] for a review-only install
+```
+
+Configure it. Every setting is an environment variable; a `.env` file in the working
+directory is read at startup:
+
+```bash
+cp .env.example .env && $EDITOR .env
+```
+
+`.env.example` lists every variable with its default and no values at all. For a
+loopback installation you need exactly one:
+
+```bash
+PHOTO_PIPELINE_LIBRARY_ROOTS=/home/you/Pictures
+```
+
+Then migrate and start the two processes:
+
+```bash
+.venv/bin/python -m photo_pipeline migrate    # create or upgrade the database
+.venv/bin/python -m photo_pipeline serve      # API and browser app on 127.0.0.1:8000
+.venv/bin/python -m photo_pipeline worker     # second terminal
+```
+
+**Both processes are required.** `serve` enqueues work and renders the application;
+nothing is scanned, scored, analysed, renamed, uploaded, or archived without a
+worker. Open <http://127.0.0.1:8000/app/> and confirm the workflow page loads.
+
+### The configuration file
+
+`.env`, or any path named by `PHOTO_PIPELINE_ENV_FILE`. It is parsed, never executed:
+`KEY=value` lines, `#` comments, optional quotes, no interpolation and no `export`. A
+configuration file that can run code is a configuration file that can be a
+vulnerability.
+
+**Anything already exported in the shell wins.** The file is your standing
+configuration; the environment is the override for one run.
+
+The archived CLI's names still work, so an existing `photo_analyzer.env` can be used
+as it is:
+
+| in the file | applied as |
+|---|---|
+| `LLM_API_KEY` / `GEMINI_API_KEY` | `OPENAI_API_KEY` |
+| `LLM_BASE_URL` | `OPENAI_BASE_URL` |
+| `LIBRARY` | `PHOTO_PIPELINE_LIBRARY_ROOTS` |
+
+`.env` and `*.env` are gitignored and refused by the repository's safety checks. The
+file holds a real key; it must never be committed.
+
+## Container installation
+
+```bash
+cp .env.example .env && $EDITOR .env
+docker compose up -d --build
+```
+
+The composition is one `serve` container, one `worker` container, a one-shot
+`migrate` that both wait for, one bind-mounted library, and one named data volume.
+
+Three variables it cannot start without:
+
+| variable | is |
+|---|---|
+| `PHOTO_PIPELINE_LIBRARY_HOST_PATH` | the library on this host |
+| `PHOTO_PIPELINE_LIBRARY_ROOTS` | where that library is mounted **inside** the container |
+| `PHOTO_PIPELINE_ACCESS_SECRET` | required, because publishing a port means the app is reachable from outside the container |
+
+Set `PHOTO_PIPELINE_UID` and `PHOTO_PIPELINE_GID` to the owner of the library: what
+the containers rename and rewrite keeps that ownership.
+
+### The data volume
+
+The `data` volume holds the database, its write-ahead log, the thumbnail cache, the
+operation journals, and the backups. **It must stay on a local filesystem.** SQLite
+in WAL mode needs real local locking, so NFS, SMB, and network volume drivers do not
+slow it down — they corrupt it. The library bind mount has no such restriction.
+
+### Ports and reaching it
+
+The API port is published to `127.0.0.1` unless `PHOTO_PIPELINE_PUBLISH_ADDRESS` says
+otherwise. Being reachable *was* the authentication in earlier versions: whoever could
+open the port owned the library. So the moment the application answers to anything but
+loopback — a hostname, a proxy, `0.0.0.0` — the access secret becomes mandatory and
+`serve` refuses to start without it rather than publishing your photographs.
+
+Behind a reverse proxy:
+
+```bash
+PHOTO_PIPELINE_ALLOWED_HOSTS=photos.example.com
+PHOTO_PIPELINE_ACCESS_SECRET=…                  # python -c 'import secrets; print(secrets.token_urlsafe(32))'
+PHOTO_PIPELINE_TRUSTED_PROXIES=10.0.0.2         # only the proxy's own address
+```
+
+`X-Forwarded-Proto` and `X-Forwarded-Host` are believed only from a trusted-proxy
+address, so a client cannot declare its own origin. The browser asks for the secret
+once per tab. Wrong secrets are rate-limited and logged with the caller's address
+only. The health endpoints stay open so an orchestrator can restart the container;
+nothing else is.
+
+### Deployment
+
+The stack is managed from git by Portainer, which redeploys on a webhook after CI
+publishes an image built from `main`. Runtime secrets live in the Portainer stack's
+environment rather than in the repository, so rotation happens in one place and
+reading the repository discloses nothing.
+
+## The settings
+
+Every variable, its default, and whether it is a secret. `.env.example` is the same
+list in copyable form.
+
+### Library and data
+
+| variable | default | meaning |
+|---|---|---|
+| `PHOTO_PIPELINE_LIBRARY_ROOTS` | none | the library boundary, `os.pathsep`-separated. No path outside these roots is ever read or written |
+| `PHOTO_PIPELINE_DATA_DIR` | `data` | database, WAL, thumbnail cache, journals, backups. Never inside the library |
+| `PHOTO_PIPELINE_DB_PATH` | `<data dir>/photo_pipeline.db` | the database file, if it must live elsewhere |
+| `PHOTO_PIPELINE_ENV_FILE` | `.env` | where to read the configuration file from |
+
+### Serving and the trust boundary
+
+| variable | default | meaning |
+|---|---|---|
+| `PHOTO_PIPELINE_HOST` | `127.0.0.1` | bind address |
+| `PHOTO_PIPELINE_PORT` | `8000` | port |
+| `PHOTO_PIPELINE_ALLOWED_HOSTS` | none | comma-separated names the app answers to besides loopback. Naming one makes the access secret mandatory |
+| `PHOTO_PIPELINE_ACCESS_SECRET` | none | **secret.** Traded for the session cookie at `GET /api/v1/session` |
+| `PHOTO_PIPELINE_TRUSTED_PROXIES` | none | comma-separated peer addresses whose forwarded headers may be believed |
+| `PHOTO_PIPELINE_MAX_REQUEST_BYTES` | `1048576` | largest request body accepted |
+
+### Logging
+
+| variable | default | meaning |
+|---|---|---|
+| `PHOTO_PIPELINE_LOG_LEVEL` | `INFO` | standard Python levels |
+| `PHOTO_PIPELINE_LOG_FORMAT` | `json` | `json` or `text` |
+
+### Limits
+
+| variable | default | meaning |
+|---|---|---|
+| `PHOTO_PIPELINE_THUMBNAIL_CACHE_QUOTA_BYTES` | `500000000` | cache quota; over it is a diagnostics warning |
+| `PHOTO_PIPELINE_THUMBNAIL_MAX_PIXELS` | `100000000` | refuse to decode anything larger. This is the decompression-bomb guard |
+| `PHOTO_PIPELINE_ARCHIVE_FREE_SPACE_RESERVE_BYTES` | `1000000000` | free space an archive destination must keep beyond the transfer itself |
+
+### The safety gate
+
+| variable | default | meaning |
+|---|---|---|
+| `PHOTO_PIPELINE_REQUIRE_DRY_RUN_APPROVAL` | `false` | refuse every mutating request until a read-only dry run of this library has been produced and approved |
+
+### External services
+
+| variable | default | meaning |
+|---|---|---|
+| `PHOTO_PIPELINE_VISION_API_KEY` | none | **secret.** Without it the analysis stage cannot run |
+| `PHOTO_PIPELINE_IMMICH_SERVER_URL` | none | the Immich server |
+| `PHOTO_PIPELINE_IMMICH_API_KEY` | none | **secret.** |
+| `PHOTO_PIPELINE_IMMICH_GO_BINARY` | `immich-go` | the uploader, found on `PATH` |
+
+### Composition only
+
+Read by `docker-compose.yml`, not by the application:
+`PHOTO_PIPELINE_IMAGE`, `PHOTO_PIPELINE_LIBRARY_HOST_PATH`,
+`PHOTO_PIPELINE_PUBLISH_ADDRESS`, `PHOTO_PIPELINE_UID`, `PHOTO_PIPELINE_GID`.
+
+A secret is never written to a log, never returned by the API, never stored in the
+database, and never recorded in a backup manifest — a manifest says `configured`, not
+the value. Do not put a real key in an example, a ticket, or a screenshot.
+
+## First run checklist
+
+Not "it started" — verified.
+
+1. **The database is at the current revision.**
+   `python -m photo_pipeline migrate` exits 0.
+2. **The application answers.**
+   `curl -s localhost:8000/api/v1/health/ready` returns 200.
+3. **The library was found.** Open the app, run a scan from the Inventory view, and
+   check the count against what you expect. Anything under `_IGNORE/` must be missing
+   from it — that is the exclusion working, not a bug.
+4. **The worker is claiming.** The scan job reaches `succeeded`. If it stays
+   `queued`, no worker is running.
+5. **Diagnostics are clean.**
+   `python -m photo_pipeline diagnostics` reports free space and an empty `warnings`.
+6. **Nothing was modified.** The scan is read-only; your files' timestamps are
+   unchanged.
+
+Only then point it at the whole library.
+
+## Operations
+
+Every operation is the same CLI, on a host or in the composition:
+
+```bash
+python -m photo_pipeline diagnostics
+docker compose run --rm --no-deps api diagnostics
+```
+
+`--no-deps` keeps a one-off command from starting a second stack; `api` is only the
+service it borrows the image and mounts from.
+
+### Upgrading
+
+1. `python -m photo_pipeline backup --reason before-upgrade`
+2. Pull the new version (or `docker compose pull && docker compose up -d`).
+3. Migrations run by themselves at startup, and a **pending schema change is
+   snapshotted first**. If the upgrade fails, the previous database and its
+   `pre-migration` backup are both intact, and the error log names the backup
+   directory.
+
+An up-to-date database is not backed up again on every start.
+
+### Backing up
+
+```bash
+python -m photo_pipeline backup --reason weekly --keep 7
+python -m photo_pipeline verify-backup data/backups/<name>
+```
+
+Backups go through SQLite's online backup API, never a file copy: with WAL enabled
+the `.db` file alone is missing every committed page still in the write-ahead log.
+Each backup is a directory holding the snapshot and a `manifest.json` — schema
+revision, SHA-256, row counts, the archive media the library depends on, and which
+settings were configured.
+
+`verify-backup` runs `PRAGMA integrity_check` **and** `PRAGMA foreign_key_check`,
+compares the snapshot's SHA-256 against the manifest, and re-counts every table it
+recorded. Bit rot, a truncated copy, and a "repaired" snapshot all fail it.
+
+`--keep N` prunes the oldest and never the newest. The same is available at
+`GET /api/v1/diagnostics`, `GET|POST /api/v1/backups`,
+`GET /api/v1/backups/{name}/verify`, and `POST /api/v1/backups/prune`.
+
+### Restoring
+
+Restore is deliberately **not** an API call. It replaces the state of an
+installation, so it belongs to a stopped one and a person at a terminal.
+
+1. Stop `serve` and `worker`.
+2. `python -m photo_pipeline verify-backup data/backups/<name>` — never restore an
+   unverified snapshot, and `restore` will refuse one anyway.
+3. `python -m photo_pipeline restore data/backups/<name> --into /path/to/fresh-data`.
+   A target that already holds a database is refused; recovering in place means
+   moving the old data directory aside first.
+4. Point `PHOTO_PIPELINE_DATA_DIR` at the restored directory and run `migrate`.
+5. Run an inventory scan, so paths are reconciled against the real library.
+6. Mount every archive location the manifest names before archiving again. The
+   database records where archived originals are; it does not contain them.
+
+Practise this against a copy before you need it.
+
+### Watching it
+
+`diagnostics` reports the database, write-ahead log, thumbnail cache, uploader
+reports, backups, and logs separately, with free space and warnings for low disk
+(`disk_low`, `disk_critical`), a cache over its quota, a write-ahead log outgrowing
+its database, and a legacy CLI writing the library.
+
+Logs go to stdout in JSON by default — `docker compose logs -f worker`, or your
+service manager's journal on a host. Uploader output is kept per attempt under the
+data directory with the API key scrubbed line by line.
+
+## When it refuses to start
+
+Every refusal below is deliberate. The application would rather stop and explain than
+guess about somebody's photographs.
+
+| exit code | meaning | what to do |
+|---|---|---|
+| `1` | the operation failed and said why — an unverifiable backup, an occupied restore target, an unknown benchmark profile | read the message; nothing was changed |
+| `2` | **the library lock is held.** Another `serve` or `worker` owns this data directory | stop the other process. The message names the holder; a lock whose process is gone is taken over automatically |
+| `3` | **a legacy CLI looks active.** The frozen command-line tools are writing this library | stop them. `--allow-legacy` overrides and you own the outcome |
+| `4` | **configuration refused.** The application is reachable beyond loopback and no access secret is set | set `PHOTO_PIPELINE_ACCESS_SECRET`, or bind to loopback only |
+| `5` | **a library root is not mounted** (containers only) | fix the bind mount. Refusing is what stops the container writing into its own throwaway layer instead of your library |
+
+Other things that look like faults and are not:
+
+- **`403 host_not_allowed`** — the `Host` header is not a name in
+  `PHOTO_PIPELINE_ALLOWED_HOSTS`. Add the real hostname; do not add `*`.
+- **`409 lock_held`** — a mutating job is already running. One at a time is what
+  makes a crash recoverable.
+- **`409 rename_recovery_required`** — an interrupted rename is unresolved. Resolve
+  it in the Renames view; unrelated mutations stay blocked until then, on purpose.
+- **A job stuck in `queued`** — no worker is running.
+- **An empty scan** — check `PHOTO_PIPELINE_LIBRARY_ROOTS`, and in a container check
+  that the path is the *container-side* mount, not the host path.
+
+## What an installer must not work around
+
+- **One writer.** Do not run two workers, and do not run the frozen CLI beside the
+  application. Each is safe alone and destructive together — the lock is not
+  bureaucracy.
+- **`_IGNORE/` is never read.** Do not "fix" the exclusion.
+- **EXIF is verified before upload.** Do not skip the checkpoint to make a stage
+  finish; an unverified projection is exactly the case where the wrong bytes reach
+  Immich.
+- **The data directory is not the library.** Keep them apart, and keep the data
+  directory on a local filesystem.
+- **Secrets stay in the environment.** Not in the database, not in a log, not in a
+  commit.

docs/overview.md

@@ -0,0 +1,77 @@
+# Overview
+
+[← Documentation index](index.md)
+
+Photo Pipeline sorts a photo library. It finds duplicates before anything expensive
+happens to them, asks a human which pictures may leave the machine, describes the
+ones that may, proposes album names from what it found, renames folders under a
+crash-safe journal, uploads through `immich-go`, and can archive a finished album off
+active storage without ever forgetting it existed.
+
+It runs locally. It talks to exactly three things outside itself: a vision provider,
+an Immich server, and `exiftool` — and it will tell you before it uses any of them.
+
+## The workflow
+
+Each stage is a gate, not a tab. A later stage can always be looked at; its actions
+stay disabled until what they depend on is true.
+
+```mermaid
+flowchart TD
+  I["0 · Inventory<br/>discover, hash, cluster duplicates"] --> S["1 · Safety<br/>score and human decision"]
+  S -->|sfw| A["2 · Analysis<br/>vision provider, EXIF checkpoint"]
+  S -->|nsfw| U
+  A --> B["3 · Albums<br/>proposal, then guarded rename"]
+  B --> U["4 · Upload<br/>immich-go, verified bytes"]
+  U --> R["5 · Archive<br/>copy, verify, then reclaim space"]
+```
+
+| Stage | What it decides | What it changes |
+|---|---|---|
+| Inventory | which file is the canonical copy of a picture | nothing — it only reads |
+| Safety | whether a photo may be sent to a cloud provider | one `sfw`/`nsfw` EXIF keyword |
+| Analysis | what a photo shows | a managed caption segment and additive keywords |
+| Albums | what a folder should be called | folder names, through a journaled rename |
+| Upload | which exact bytes reach Immich | nothing locally; assets appear in Immich |
+| Archive | which album leaves active storage | files move to the archive, after verification |
+
+## What it will not do
+
+These are enforced in code, not by convention, and each one is why some action you
+expected is sometimes refused.
+
+- **Nothing under `_IGNORE/` is ever read.** Not scanned, not counted, not
+  thumbnailed, not sent anywhere.
+- **A path is not an identity.** Every picture has a stable id, so moving or renaming
+  it loses no history.
+- **Only a confirmed-SFW photo reaches the vision provider.** A photo marked NSFW is
+  still uploadable to Immich; it simply never leaves for analysis.
+- **Metadata is verified, not hoped for.** Every stage that writes EXIF reads it back
+  and proves that what it did not own is unchanged.
+- **One writer at a time.** A library lock, held by one process, is what makes a
+  crash recoverable instead of ambiguous.
+- **Nothing irreversible happens without a preview and an explicit approval** that
+  names the exact count.
+
+## Where things live
+
+| | |
+|---|---|
+| the photo library | wherever you point `PHOTO_PIPELINE_LIBRARY_ROOTS`; mounted read-write |
+| the database, cache, logs, backups | the data directory, never inside the library |
+| secrets | the environment, never the database and never a log line |
+| these documents | `docs/` in the repository, served at `/docs` by the application |
+
+## Running it
+
+The short version, for a host installation:
+
+```bash
+python -m photo_pipeline migrate
+python -m photo_pipeline serve      # the API and this browser application
+python -m photo_pipeline worker     # the process that does the long work
+```
+
+The full procedure, the container path, and every setting belong to the installation
+manual, which is being written next — see
+[the documentation index](index.md#being-written).

frontend/css/app.css

@@ -170,6 +170,20 @@ a.link:hover { text-decoration: underline; }
 .grid th, .grid td { text-align: left; padding: 6px 8px; border-bottom: 1px solid var(--border); vertical-align: top; }
 .path { font-family: ui-monospace, monospace; font-size: 0.85em; word-break: break-all; }
 
+/* Docs: prose, so it gets a reading measure rather than the full window width. */
+.doc { max-width: 72ch; }
+.doc h1 { margin-top: 0; }
+.doc h2, .doc h3 { margin-top: 1.6em; border-bottom: 1px solid var(--border); padding-bottom: 4px; }
+.doc code { font-family: ui-monospace, monospace; font-size: 0.9em; background: var(--surface-2); padding: 1px 4px; border-radius: 4px; }
+.doc pre { background: var(--surface-2); border: 1px solid var(--border); border-radius: var(--radius); padding: 12px; overflow: auto; }
+.doc pre code { background: none; padding: 0; }
+.doc table { border-collapse: collapse; width: 100%; }
+.doc th, .doc td { text-align: left; padding: 6px 8px; border-bottom: 1px solid var(--border); vertical-align: top; }
+.doc blockquote { margin: 1em 0; padding-left: 12px; border-left: 3px solid var(--border); color: var(--muted); }
+.doc img { max-width: 100%; }
+.diagram { margin: 1.5em 0; padding: 0; overflow: auto; }
+.diagram svg { max-width: 100%; height: auto; }
+
 /* Narrow screens: stack the panes so blockers and actions stay reachable. */
 @media (max-width: 720px) {
   .two-pane { grid-template-columns: 1fr; }

frontend/index.html

@@ -21,6 +21,7 @@
         <a href="#/uploads" data-nav="uploads">Upload</a>
         <a href="#/archive" data-nav="archive">Archive</a>
         <a href="#/stats" data-nav="stats">Stats</a>
+        <a href="#/docs" data-nav="docs">Docs</a>
       </nav>
     </header>
     <main id="app" aria-live="polite"><!-- views render here --></main>

frontend/js/app.js

@@ -1,5 +1,6 @@
 import { api } from "./api.js";
 import { renderArchive, setArchiveRender } from "./archive.js";
+import { renderDocs } from "./docs.js";
 import { navigate, onRouteChange, parseHash } from "./router.js";
 import { renderRenames, setRenamesRender } from "./renames.js";
 import { renderUploads, setUploadsRender } from "./uploads.js";
@@ -385,6 +386,7 @@ function render() {
   else if (path === "/uploads") renderUploads(root, params);
   else if (path === "/archive") renderArchive(root, params);
   else if (path === "/stats") renderStats(root, params);
+  else if (path === "/docs") renderDocs(root, params);
   else show(errorBanner("Unknown view"));
 }
 

frontend/js/docs.js

@@ -0,0 +1,263 @@
+// The documentation view (US09-01): the manuals in `docs/`, rendered in the app.
+//
+// The same markdown files are the repository's documentation and the deployment's
+// documentation. An operator who was handed a URL and an access secret has no
+// checkout in front of them, and the network the application runs on is not assumed
+// to reach a CDN — so the renderer is vendored and everything here is same-origin.
+//
+// Nothing on this page is authenticated. It is served from the static mount beside
+// the application shell, exactly like `index.html`: the troubleshooting page is
+// needed most by whoever cannot get past the access secret, and no documentation
+// file contains anything a session would protect.
+//
+// `marked` and `mermaid` are loaded lazily, on the first documentation page and the
+// first diagram. They are large, and every other view does without them.
+import { el, errorBanner, setActiveNav } from "./dom.js";
+
+const DOCS_BASE = "/docs";
+const VENDOR = "/app/js/vendor";
+const INDEX_PAGE = "index";
+// A page name comes from the hash, so it is caller input: no scheme, no traversal,
+// no absolute path. The server would refuse those too; this refuses them earlier and
+// without a request that looks like an attempt.
+const PAGE_PATTERN = /^[\w-]+(\/[\w-]+)*$/;
+
+let markedPromise;
+let mermaidPromise;
+
+function loadMarked() {
+  markedPromise ??= import(`${VENDOR}/marked.esm.js`).then((module) => module.marked);
+  return markedPromise;
+}
+
+// mermaid ships one large UMD bundle rather than a self-contained ES module, so it
+// arrives through a script element. `script-src 'self'` allows it because it is ours
+// and same-origin; nothing here relaxes that.
+function loadMermaid() {
+  mermaidPromise ??= new Promise((resolve, reject) => {
+    const script = document.createElement("script");
+    script.src = `${VENDOR}/mermaid.min.js`;
+    script.onload = () => resolve(window.mermaid);
+    script.onerror = () => reject(new Error("the diagram renderer could not be loaded"));
+    document.head.appendChild(script);
+  });
+  return mermaidPromise;
+}
+
+// ── the pure parts, unit-tested in js/tests/unit.js ──────────────────────────
+
+/** A stable, readable heading anchor. Letters and digits of any script survive. */
+export function slug(text) {
+  const cleaned = String(text)
+    .toLowerCase()
+    .trim()
+    .replace(/[^\p{L}\p{N}\s-]/gu, "")
+    .replace(/[\s-]+/g, "-")
+    .replace(/^-|-$/g, "");
+  return cleaned || "section";
+}
+
+/** Assign every heading an id, disambiguating repeats the way a reader would
+ *  expect: the first `Notes` keeps `#notes`, the second becomes `#notes-2`. */
+export function assignHeadingIds(headings) {
+  const used = new Map();
+  for (const heading of headings) {
+    const base = slug(heading.textContent);
+    const seen = (used.get(base) || 0) + 1;
+    used.set(base, seen);
+    heading.id = seen === 1 ? base : `${base}-${seen}`;
+  }
+  return headings;
+}
+
+/**
+ * Where a link inside a documentation page should go.
+ *
+ * Markdown links between documents are relative file paths, which a browser would
+ * treat as downloads that leave the application. Returns the in-app target for a
+ * link to another document or to a heading, and `null` for anything else — external
+ * links stay exactly as the author wrote them.
+ */
+export function resolveDocLink(currentPage, href) {
+  if (!href) return null;
+  if (/^[a-z][a-z\d+.-]*:/i.test(href) || href.startsWith("//")) return null;
+  if (href.startsWith("#")) return { page: currentPage, anchor: href.slice(1) };
+  const [target, anchor = ""] = href.split("#");
+  if (!/\.md$/i.test(target)) return null;
+  // Resolved against the current document, so `../guides/x.md` means what it means
+  // in the repository. The origin is a placeholder; only the path is used.
+  const resolved = new URL(target, `https://docs.invalid/${currentPage}.md`);
+  const page = decodeURIComponent(resolved.pathname).replace(/^\//, "").replace(/\.md$/i, "");
+  return PAGE_PATTERN.test(page) ? { page, anchor } : null;
+}
+
+/**
+ * Point every in-app link at its route, and leave every other link alone.
+ *
+ * The resolved page is kept on the element, so the reading order can be read back
+ * from the rendered index rather than parsed out of the markdown a second time.
+ */
+export function rewriteLinks(article, page) {
+  for (const link of article.querySelectorAll("a[href]")) {
+    const href = link.getAttribute("href");
+    const target = resolveDocLink(page, href);
+    if (target) {
+      link.setAttribute("href", docHash(target.page, target.anchor));
+      link.dataset.docPage = target.page;
+    } else if (/^https?:/i.test(href)) {
+      link.setAttribute("rel", "noreferrer noopener");
+      link.setAttribute("target", "_blank");
+    }
+  }
+  return article;
+}
+
+/** The reading order, taken from the index document itself — one list, in one
+ *  place, that is equally the index on Gitea and the navigation here. */
+export function documentIndex(indexElement) {
+  const pages = [];
+  for (const link of indexElement.querySelectorAll("a[data-doc-page]")) {
+    const page = link.dataset.docPage;
+    if (page !== INDEX_PAGE && !pages.some((entry) => entry.page === page)) {
+      pages.push({ page, title: link.textContent.trim() });
+    }
+  }
+  return pages;
+}
+
+// ── the view ─────────────────────────────────────────────────────────────────
+
+async function fetchPage(page) {
+  const response = await fetch(`${DOCS_BASE}/${page}.md`, { headers: { accept: "text/markdown" } });
+  if (!response.ok) throw new Error(`documentation page unavailable: ${response.status}`);
+  return response.text();
+}
+
+async function toArticle(markdown, page) {
+  const marked = await loadMarked();
+  const article = el("article", { class: "doc", "data-testid": "doc", "data-page": page });
+  // The only innerHTML in the application, and deliberate: rendering markdown *is*
+  // producing HTML. The input is a file from this repository, and the page's CSP
+  // (`script-src 'self'`, no `'unsafe-inline'`) means injected script and inline
+  // handlers do not run even if one ever were not.
+  article.innerHTML = marked.parse(markdown, { async: false });
+  assignHeadingIds(article.querySelectorAll("h1, h2, h3, h4, h5, h6"));
+  rewriteLinks(article, page);
+  return article;
+}
+
+function docHash(page, anchor = "") {
+  const query = new URLSearchParams(anchor ? { page, anchor } : { page });
+  return `#/docs?${query}`;
+}
+
+/** Diagrams are authored as ```mermaid blocks so the source is diffable and Gitea
+ *  renders them natively. A failure here replaces the diagram, never the page. */
+async function renderDiagrams(article) {
+  const blocks = [...article.querySelectorAll("pre > code.language-mermaid")];
+  if (!blocks.length) return;
+  let mermaid;
+  try {
+    mermaid = await loadMermaid();
+    mermaid.initialize({ startOnLoad: false, securityLevel: "strict", theme: "dark" });
+  } catch (error) {
+    blocks.forEach((block) => block.closest("pre").replaceWith(errorBanner(error.message)));
+    return;
+  }
+  for (const [index, block] of blocks.entries()) {
+    const figure = el("figure", { class: "diagram", "data-testid": "diagram" });
+    block.closest("pre").replaceWith(figure);
+    try {
+      const { svg } = await mermaid.render(`diagram-${index}-${Date.now()}`, block.textContent);
+      figure.innerHTML = svg;
+    } catch (error) {
+      figure.replaceWith(errorBanner(`Diagram could not be drawn: ${error.message}`));
+    }
+  }
+}
+
+function notFound() {
+  return el(
+    "div",
+    { class: "alert", role: "alert", "data-testid": "doc-not-found" },
+    "That documentation page does not exist. ",
+    el("a", { class: "link", href: docHash(INDEX_PAGE) }, "Back to the documentation index")
+  );
+}
+
+function sidebar(pages, current) {
+  return el(
+    "nav",
+    { class: "card", "aria-label": "Documentation" },
+    el("h2", {}, "Documentation"),
+    el(
+      "ul",
+      { class: "album-list", "data-testid": "doc-pages" },
+      el(
+        "li",
+        {},
+        el(
+          "a",
+          {
+            href: docHash(INDEX_PAGE),
+            "aria-current": current === INDEX_PAGE ? "true" : false,
+          },
+          "Index"
+        )
+      ),
+      ...pages.map(({ page, title }) =>
+        el(
+          "li",
+          {},
+          el(
+            "a",
+            {
+              href: docHash(page),
+              "data-page": page,
+              "aria-current": page === current ? "true" : false,
+            },
+            title
+          )
+        )
+      )
+    )
+  );
+}
+
+export async function renderDocs(root, params = {}) {
+  setActiveNav("docs");
+  const requested = params.page || INDEX_PAGE;
+  const page = PAGE_PATTERN.test(requested) ? requested : "";
+
+  let indexArticle;
+  try {
+    indexArticle = await toArticle(await fetchPage(INDEX_PAGE), INDEX_PAGE);
+  } catch (error) {
+    root.replaceChildren(errorBanner(`Documentation is unavailable: ${error.message}`));
+    return;
+  }
+  const pages = documentIndex(indexArticle);
+
+  let article;
+  if (!page) article = notFound();
+  else if (page === INDEX_PAGE) article = indexArticle;
+  else {
+    try {
+      article = await toArticle(await fetchPage(page), page);
+    } catch {
+      article = notFound();
+    }
+  }
+
+  root.replaceChildren(el("div", { class: "two-pane" }, sidebar(pages, page), article));
+  await renderDiagrams(article);
+  scrollToAnchor(params.anchor);
+}
+
+// A documentation anchor cannot live in the hash — the hash is the route — so it
+// travels as a parameter and is applied after the page renders.
+function scrollToAnchor(anchor) {
+  if (!anchor) return;
+  const target = document.getElementById(anchor);
+  if (target) target.scrollIntoView({ block: "start" });
+}

frontend/js/tests/unit.js

@@ -7,6 +7,7 @@ import { createStore } from "../store.js";
 import { parseHash, navigate } from "../router.js";
 import { api, cancellable } from "../api.js";
 import { subscribeJob } from "../events.js";
+import { assignHeadingIds, documentIndex, resolveDocLink, rewriteLinks, slug } from "../docs.js";
 
 const cases = [];
 function ok(name, cond) {
@@ -142,6 +143,62 @@ async function run() {
     window.EventSource = realES;
   }
 
+  // ── documentation view (US09-01) ─────────────────────────────────────────
+  {
+    ok("slug lowercases and dashes a heading", slug("Key Flows") === "key-flows");
+    ok(
+      "slug drops punctuation but keeps words",
+      slug("What it *will not* do:") === "what-it-will-not-do"
+    );
+    ok("slug keeps non-ASCII letters", slug("Größe & Gewicht") === "größe-gewicht");
+    ok("slug never yields an empty anchor", slug("!!!") === "section");
+
+    const doc = document.createElement("div");
+    doc.innerHTML = "<h2>Notes</h2><h3>Notes</h3><h2>Notes</h2>";
+    const ids = [...assignHeadingIds(doc.querySelectorAll("h2, h3"))].map((h) => h.id);
+    ok("repeated headings get distinct anchors", ids.join(",") === "notes,notes-2,notes-3");
+
+    ok(
+      "a sibling document link becomes an in-app route",
+      resolveDocLink("index", "overview.md")?.page === "overview"
+    );
+    const nested = resolveDocLink("guides/install", "../overview.md#running-it");
+    ok("a relative link resolves against the current page", nested?.page === "overview");
+    ok("a link's anchor survives the rewrite", nested?.anchor === "running-it");
+    ok(
+      "a bare anchor stays on the current page",
+      resolveDocLink("overview", "#the-workflow")?.page === "overview"
+    );
+    ok("an external link is left alone", resolveDocLink("index", "https://example.test/x") === null);
+    ok("a non-markdown relative link is left alone", resolveDocLink("index", "images/a.png") === null);
+    // A climb cannot leave the documentation tree: it is clamped at the root, so the
+    // page it names is still fetched from under /docs and simply does not exist.
+    ok(
+      "a link that climbs above the docs root is clamped",
+      resolveDocLink("index", "../../etc/passwd.md")?.page === "etc/passwd"
+    );
+    ok(
+      "a page name that is not a page name is refused",
+      resolveDocLink("index", "..%2f..%2fetc%2fpasswd.md") === null
+    );
+
+    const index = document.createElement("div");
+    index.innerHTML =
+      '<a href="overview.md">Overview</a><a href="overview.md">again</a>' +
+      '<a href="index.md">itself</a><a href="https://example.test">out</a>';
+    const listed = documentIndex(rewriteLinks(index, "index"));
+    ok(
+      "an external link is marked safe to open away from the app",
+      index.querySelector('a[href^="https"]').rel === "noreferrer noopener"
+    );
+    ok(
+      "an in-app link points at the docs route",
+      index.querySelector("a").getAttribute("href") === "#/docs?page=overview"
+    );
+    ok("the index lists each page once, in order", listed.length === 1);
+    ok("the index takes its titles from the link text", listed[0].title === "Overview");
+  }
+
   await tick();
   const failed = cases.filter((c) => !c.ok);
   window.__RESULTS__ = { passed: cases.length - failed.length, failed: failed.length, cases };

frontend/js/vendor/VERSIONS.json

@@ -0,0 +1,23 @@
+{
+  "_comment": "Vendored browser libraries (US09-01). Committed rather than fetched: the application is deployed to a network whose outbound access is not assumed, and `default-src 'self'` forbids a CDN. Checksums are asserted by tests/integration/test_documentation.py, so replacing a file without recording it here fails the suite. Update by downloading the pinned URL and recording the new version and sha256 in the same commit.",
+  "libraries": [
+    {
+      "file": "marked.esm.js",
+      "name": "marked",
+      "version": "18.0.10",
+      "license": "MIT",
+      "url": "https://cdn.jsdelivr.net/npm/marked@18.0.10/lib/marked.esm.js",
+      "sha256": "4cf47dfebb7f614a08fc0a579ab0fe407ff0ed2b717bf953040c85b2f493a4f0",
+      "why": "Markdown to HTML for the documentation view. An ES module with no dependencies, imported lazily by js/docs.js."
+    },
+    {
+      "file": "mermaid.min.js",
+      "name": "mermaid",
+      "version": "11.17.0",
+      "license": "MIT",
+      "url": "https://cdn.jsdelivr.net/npm/mermaid@11.17.0/dist/mermaid.min.js",
+      "sha256": "8d8e0eec56d3a83b4b3c87f42050845546dee93ebe1875d2117c12e6947c0cb3",
+      "why": "Renders ```mermaid blocks so diagrams are diffable source that Gitea also renders. The single-file UMD build rather than the ES module entry, whose ~40 lazy chunks would each need vendoring and pinning."
+    }
+  ]
+}

photo_pipeline/__main__.py

@@ -1,5 +1,5 @@
 """Application management CLI:
-``python -m photo_pipeline {serve,migrate,worker,import-legacy-scores,backup,verify-backup,restore,diagnostics}``.
+``python -m photo_pipeline {serve,migrate,worker,import-legacy-scores,backup,verify-backup,restore,diagnostics,benchmark,release-gate,container-gate,dry-run,approve-dry-run}``.
 
 ``serve`` and ``worker`` take the library process lock for their role (US07-05):
 two workers, or the frozen CLI running beside the app, would each be safe on their
@@ -70,6 +70,14 @@ def main(argv: Sequence[str] | None = None) -> int:
         "release-gate", help="Run every suite in an isolated stack and keep the evidence"
     )
     gate_cmd.add_argument("--output", help="Evidence directory (default: data/release/<stamp>)")
+    container_cmd = commands.add_parser(
+        "container-gate",
+        help="Provision the composition from the built image, run the phase_h "
+        "acceptance suite against it, destroy it, and keep the evidence (US08-05)",
+    )
+    container_cmd.add_argument(
+        "--output", help="Evidence directory (default: data/container-gate/<stamp>)"
+    )
     dry_cmd = commands.add_parser(
         "dry-run", help="Read-only reconciliation of the configured library (US07-07)"
     )
@@ -155,6 +163,31 @@ def main(argv: Sequence[str] | None = None) -> int:
         )
         return 0 if report["ok"] else 1
 
+    if args.command == "container-gate":
+        from datetime import datetime, timezone
+
+        from photo_pipeline.services import release
+
+        # The suite provisions and destroys the composition itself; what this command
+        # adds is the single entry point and the retained evidence. No skip is an
+        # environment limit here: a gate that did not reach the containers proved
+        # nothing about them.
+        output = args.output or Path(config.data_dir) / "container-gate" / datetime.now(
+            timezone.utc
+        ).strftime("%Y%m%dT%H%M%SZ")
+        report = release.run_gate(
+            config,
+            output=output,
+            stages=release.CONTAINER_STAGES,
+            allowed_skips=release.CONTAINER_ALLOWED_SKIP_REASONS,
+        )
+        print(
+            json.dumps(
+                {k: v for k, v in report.items() if k not in ("stages", "matrix")}, indent=2
+            )
+        )
+        return 0 if report["ok"] else 1
+
     if args.command == "dry-run":
         from photo_pipeline.services import release
 

photo_pipeline/api/app.py

@@ -53,6 +53,7 @@ from photo_pipeline.services.thumbnails import ThumbnailService
 from photo_pipeline.services.upload_batches import UploadBatchService
 
 FRONTEND_DIR = Path(__file__).resolve().parents[2] / "frontend"
+DOCS_DIR = Path(__file__).resolve().parents[2] / "docs"
 
 log = logging.getLogger(__name__)
 
@@ -151,4 +152,10 @@ def create_app(config: Config | None = None) -> FastAPI:
     # Static single-page app (hash-routed). Mounted last so /api/v1 wins.
     if FRONTEND_DIR.is_dir():
         app.mount("/app", StaticFiles(directory=FRONTEND_DIR, html=True), name="app")
+    # The manuals, as their own markdown files (US09-01). Unauthenticated for the
+    # same reason the application shell is: whoever cannot get past the access
+    # secret is exactly who needs the troubleshooting page, and no document holds
+    # anything a session would protect.
+    if DOCS_DIR.is_dir():
+        app.mount("/docs", StaticFiles(directory=DOCS_DIR), name="docs")
     return app

photo_pipeline/api/security.py

@@ -57,9 +57,21 @@ LOOPBACK_HOSTS = frozenset({"127.0.0.1", "localhost", "::1", "[::1]"})
 # backup a careful operator takes first, and its retention (US07-07).
 MUTATION_EXEMPT_PATHS = frozenset({f"{API_PREFIX}/backups", f"{API_PREFIX}/backups/prune"})
 
-# Applied to every response. No inline script/style is used by the frontend, so the
-# policy can stay strict; `frame-ancestors 'none'` and CORP keep other pages from
+# Applied to every response. `frame-ancestors 'none'` and CORP keep other pages from
 # embedding the app or its thumbnails.
+#
+# `script-src 'self'` is the boundary that matters and it is unchanged: no
+# `'unsafe-eval'`, no `'unsafe-inline'`, so nothing injected into the DOM can execute.
+# Whether the vendored diagram renderer needed `'unsafe-eval'` was measured rather
+# than assumed — mermaid's bundle contains no `eval(` and no `new Function`, and it
+# renders under this exact policy without a single script-src violation.
+#
+# `style-src` does gain `'unsafe-inline'` (US09-01): mermaid styles the SVG it builds
+# with an injected `<style>` element and `style=` attributes, and a diagram's CSS
+# cannot be hashed in advance. The concession is bounded by the directives around it
+# — with script execution still refused and `img-src`, `connect-src`, and `font-src`
+# all `'self'`, the CSS exfiltration channels stay closed and what is left is
+# defacement of a page its own operator is already looking at.
 DEFAULT_HEADERS = {
     "x-content-type-options": "nosniff",
     "x-frame-options": "DENY",
@@ -67,8 +79,9 @@ DEFAULT_HEADERS = {
     "cross-origin-resource-policy": "same-origin",
     "cross-origin-opener-policy": "same-origin",
     "content-security-policy": (
-        "default-src 'self'; img-src 'self' data:; style-src 'self'; script-src 'self'; "
-        "connect-src 'self'; frame-ancestors 'none'; base-uri 'none'; form-action 'none'"
+        "default-src 'self'; img-src 'self' data:; style-src 'self' 'unsafe-inline'; "
+        "script-src 'self'; connect-src 'self'; font-src 'self'; "
+        "frame-ancestors 'none'; base-uri 'none'; form-action 'none'"
     ),
 }
 

photo_pipeline/services/release.py

@@ -60,6 +60,17 @@ ALLOWED_SKIP_REASONS = (
     "bind-mount ownership is virtualised",
 )
 
+# The container acceptance gate (US08-05): the deployed application, verified the way
+# the host application is. Deliberately one stage selected by marker, so adding a
+# phase_h test is enough to put it in front of a deploy.
+CONTAINER_STAGES: tuple[tuple[str, tuple[str, ...]], ...] = (
+    ("container", ("tests/e2e", "-m", "phase_h")),
+)
+
+# Nothing. This gate exists to prove the *deployed* runtime, and every reason a check
+# would skip here — no daemon, no compose, no browser — means it was not proven.
+CONTAINER_ALLOWED_SKIP_REASONS: tuple[str, ...] = ()
+
 
 class ReleaseError(RuntimeError):
     pass
@@ -175,13 +186,15 @@ def _run_stage(label: str, paths: tuple[str, ...], *, repo: Path, log_dir: Path)
     return StageResult(label, command, result.returncode, elapsed, summary, skipped)
 
 
-def unexpected_skips(results: list[StageResult]) -> list[str]:
+def unexpected_skips(
+    results: list[StageResult], allowed: tuple[str, ...] = ALLOWED_SKIP_REASONS
+) -> list[str]:
     """Skips the gate will not accept: everything but the documented environment ones."""
     return [
         line
         for result in results
         for line in result.skipped
-        if not any(reason in line for reason in ALLOWED_SKIP_REASONS)
+        if not any(reason in line for reason in allowed)
     ]
 
 
@@ -190,6 +203,7 @@ def run_gate(
     *,
     output: Path | str | None = None,
     stages: tuple[tuple[str, tuple[str, ...]], ...] = STAGES,
+    allowed_skips: tuple[str, ...] = ALLOWED_SKIP_REASONS,
     repo: Path | None = None,
 ) -> dict:
     """Run every suite in an isolated stack and retain checksummed evidence.
@@ -206,13 +220,14 @@ def run_gate(
     logs = directory / "logs"
     logs.mkdir(parents=True, exist_ok=True)
 
+    started_at = _now()
     matrix = story_matrix(repo)
     results = [_run_stage(label, paths, repo=repo, log_dir=logs) for label, paths in stages]
-    skips = unexpected_skips(results)
+    skips = unexpected_skips(results, allowed_skips)
 
     report = {
         "schema_version": SCHEMA_VERSION,
-        "started_at": _now().isoformat(),
+        "started_at": started_at.isoformat(),
         "revision": revision(),
         "python": sys.version.split()[0],
         "platform": os.uname().sysname,

pyproject.toml

@@ -53,4 +53,5 @@ markers = [
     "phase_e: Phase E end-to-end acceptance (US05-06) — upload preflight, uploader, and browser journeys",
     "phase_f: Phase F end-to-end acceptance (US06-06) — archive destination, transfer, and restore journeys",
     "container: builds and runs the container image and its composition (US08-02, US08-03) — needs a Docker daemon, the compose plugin, and the network",
+    "phase_h: Phase H container deployment acceptance (US08-05) — the browser, upgrade, restart, and security journeys against the composed stack",
 ]

tests/e2e/_container_harness.py

@@ -0,0 +1,260 @@
+"""The composed stack, as a fixture: build, provision, drive, destroy (US08-03/US08-05).
+
+Extracted from ``tests/e2e/test_compose_stack.py`` when the container acceptance gate
+needed the same stack under a different image, project, and library. One
+implementation, because two would drift on exactly the details that make a container
+test worth anything — the mount, the volume, the ports, and the teardown.
+
+Nothing here fakes anything below the process boundary: Docker builds the image,
+Compose starts the real containers, and every helper talks to them over HTTP or the
+Docker CLI.
+"""
+
+from __future__ import annotations
+
+import os
+import shutil
+import socket
+import subprocess
+import tempfile
+import time
+from pathlib import Path
+
+import httpx
+import pytest
+from PIL import Image
+
+REPO = Path(__file__).resolve().parents[2]
+COMPOSE_FILE = REPO / "docker-compose.yml"
+CONTAINER_LIBRARY = "/library"
+READY_TIMEOUT_SECONDS = 180
+JOB_TIMEOUT_SECONDS = 300
+UP_TIMEOUT_SECONDS = 30 * 60
+
+
+def compose_available() -> bool:
+    try:
+        return (
+            subprocess.run(
+                ["docker", "compose", "version"], capture_output=True, timeout=60
+            ).returncode
+            == 0
+        )
+    except (OSError, subprocess.SubprocessError):
+        return False
+
+
+def free_port() -> int:
+    with socket.socket() as sock:
+        sock.bind(("127.0.0.1", 0))
+        return sock.getsockname()[1]
+
+
+def mount_base() -> Path:
+    """A directory a Docker VM shares with the host.
+
+    Not pytest's ``tmp_path``: on macOS that is ``/var/folders/...``, which a Docker VM
+    (Colima, Docker Desktop) does not share, so the bind mount would arrive empty and
+    every assertion would be about nothing. ``$HOME`` is shared by every default
+    configuration.
+    """
+    base = Path(
+        os.environ.get("PHOTO_PIPELINE_TEST_MOUNT_BASE", Path.home() / ".cache" / "photo-pipeline")
+    )
+    base.mkdir(parents=True, exist_ok=True)
+    return base
+
+
+def temporary_library(albums: dict[str, int], *, prefix: str = "library-") -> Path:
+    """A fixture library on a shareable host path, with the exclusion sentinel in it."""
+    root = Path(tempfile.mkdtemp(prefix=prefix, dir=mount_base()))
+    for position, (album, count) in enumerate(albums.items()):
+        (root / album).mkdir(parents=True, exist_ok=True)
+        for index in range(count):
+            # Distinct per album *and* index: two solid images of the same colour are
+            # byte-identical, which would make them a duplicate cluster by accident.
+            colour = (17 + 7 * index, 31 + 29 * position, 160 - 3 * index)
+            Image.new("RGB", (64, 48), colour).save(root / album / f"{album}_{index}.jpg")
+    # Never discovered, counted, analyzed, or uploaded — asserted from outside it.
+    (root / "_IGNORE").mkdir(exist_ok=True)
+    Image.new("RGB", (32, 32), (0, 0, 0)).save(root / "_IGNORE" / "sentinel.jpg")
+    return root
+
+
+def remove_library(root: Path) -> None:
+    shutil.rmtree(root, ignore_errors=True)
+
+
+def write_env_file(path: Path, secret: str, extra: dict[str, str] | None = None) -> Path:
+    """Configuration and secrets come from the environment, so a test writes its own
+    file rather than borrowing the operator's ``.env``."""
+    lines = [
+        f"PHOTO_PIPELINE_ACCESS_SECRET={secret}",
+        "PHOTO_PIPELINE_LOG_FORMAT=text",
+        # The deterministic vision seam, in the data volume so both roles and the test
+        # can read it (concept §18).
+        "PHOTO_PIPELINE_FAKE_VISION_LOG=/data/vision.log",
+        *(f"{key}={value}" for key, value in (extra or {}).items()),
+    ]
+    path.write_text("\n".join(lines) + "\n")
+    return path
+
+
+class Stack:
+    """The composition under test, plus the environment it was started with."""
+
+    def __init__(
+        self,
+        library: Path,
+        env_file: Path,
+        *,
+        project: str,
+        image: str,
+        secret: str,
+    ) -> None:
+        self.library = library
+        self.project = project
+        self.secret = secret
+        self.port = free_port()
+        self.base = f"http://127.0.0.1:{self.port}"
+        # Compose reads the repository's own .env for substitution; the process
+        # environment wins over it, so the test's values are the ones that apply.
+        self.env = {
+            **os.environ,
+            "PHOTO_PIPELINE_IMAGE": image,
+            "PHOTO_PIPELINE_ENV_FILE": str(env_file),
+            "PHOTO_PIPELINE_LIBRARY_HOST_PATH": str(library),
+            "PHOTO_PIPELINE_LIBRARY_ROOTS": CONTAINER_LIBRARY,
+            "PHOTO_PIPELINE_PORT": str(self.port),
+            "PHOTO_PIPELINE_UID": str(os.getuid()),
+            "PHOTO_PIPELINE_GID": str(os.getgid()),
+        }
+
+    # ── the compose lifecycle ────────────────────────────────────────────────
+
+    def compose(self, *args: str, check: bool = True, timeout: int = 300):
+        result = subprocess.run(
+            ["docker", "compose", "-p", self.project, "-f", str(COMPOSE_FILE), *args],
+            capture_output=True,
+            text=True,
+            env=self.env,
+            cwd=REPO,
+            timeout=timeout,
+        )
+        if check and result.returncode != 0:
+            raise AssertionError(
+                f"docker compose {' '.join(args)} failed:\n{result.stdout}\n{result.stderr}\n"
+                f"{self.compose('logs', '--tail', '80', check=False).stdout}"
+            )
+        return result
+
+    def up(self, *extra: str) -> None:
+        self.compose("up", "--detach", *extra, timeout=UP_TIMEOUT_SECONDS)
+
+    def down(self, *, volumes: bool = True) -> None:
+        self.compose(
+            "down",
+            *(("--volumes",) if volumes else ()),
+            "--remove-orphans",
+            check=False,
+            timeout=300,
+        )
+
+    def use_image(self, image: str) -> None:
+        """Point the composition at another tag — the upgrade path (US08-05)."""
+        self.env["PHOTO_PIPELINE_IMAGE"] = image
+
+    def logs(self, *services: str) -> str:
+        result = self.compose("logs", *services, check=False)
+        return result.stdout + result.stderr
+
+    def wait_until_ready(self) -> None:
+        deadline = time.monotonic() + READY_TIMEOUT_SECONDS
+        while time.monotonic() < deadline:
+            try:
+                if httpx.get(f"{self.base}/api/v1/health/ready", timeout=5).status_code == 200:
+                    return
+            except httpx.HTTPError:
+                pass
+            time.sleep(0.5)
+        raise AssertionError(f"the stack never became ready:\n{self.logs()}")
+
+    # ── talking to it ────────────────────────────────────────────────────────
+
+    def client(self) -> httpx.Client:
+        """A browser that has loaded the app: session cookie in the jar, token in a
+        header. The session belongs to the API process, so it is re-bootstrapped after
+        every restart."""
+        client = httpx.Client(base_url=f"{self.base}/api/v1", timeout=60)
+        bootstrap = client.get("/session", headers={"X-Access-Secret": self.secret})
+        assert bootstrap.status_code == 200, bootstrap.text
+        client.headers["X-CSRF-Token"] = bootstrap.json()["csrf_token"]
+        return client
+
+
+def await_job(client: httpx.Client, job_id: str, states=("succeeded",)) -> dict:
+    deadline = time.monotonic() + JOB_TIMEOUT_SECONDS
+    snapshot: dict = {}
+    while time.monotonic() < deadline:
+        response = client.get(f"/jobs/{job_id}")
+        if response.status_code == 200:
+            snapshot = response.json()
+            if snapshot["state"] in states:
+                return snapshot
+        time.sleep(0.5)
+    raise AssertionError(f"job {job_id} never reached {states}: {snapshot}")
+
+
+def build_image(tag: str, *, revision: str | None = None) -> str:
+    """Build the application image; from a git revision's tree when one is named.
+
+    ``revision`` is how the upgrade journey gets the *previous* version without a
+    registry: the tree of that commit is the build context, so what it produces is the
+    image that commit would have published.
+    """
+    if revision is None:
+        subprocess.run(
+            [
+                "docker",
+                "build",
+                "--build-arg",
+                f"UID={os.getuid()}",
+                "--build-arg",
+                f"GID={os.getgid()}",
+                "-t",
+                tag,
+                str(REPO),
+            ],
+            check=True,
+            timeout=UP_TIMEOUT_SECONDS,
+        )
+        return tag
+    archive = subprocess.run(
+        ["git", "archive", "--format=tar", revision],
+        cwd=REPO,
+        capture_output=True,
+        check=True,
+        timeout=300,
+    ).stdout
+    subprocess.run(
+        [
+            "docker",
+            "build",
+            "--build-arg",
+            f"UID={os.getuid()}",
+            "--build-arg",
+            f"GID={os.getgid()}",
+            "-t",
+            tag,
+            "-",
+        ],
+        input=archive,
+        check=True,
+        timeout=UP_TIMEOUT_SECONDS,
+    )
+    return tag
+
+
+needs_compose = pytest.mark.skipif(
+    not compose_available(), reason="no Docker daemon with the compose plugin"
+)

tests/e2e/test_compose_stack.py

@@ -10,188 +10,63 @@ what this story is about — the mount, the lock, the volume, and the restart ar
 The file contract (one API, one worker, migrations first, no committed values) is
 checked without a daemon in ``tests/integration/test_compose_runtime.py``; only the
 running proof needs Docker, and CI is where it runs unskipped (US08-04).
+
+The stack itself lives in ``tests/e2e/_container_harness.py``, shared with the
+container acceptance gate (US08-05).
 """
 
 from __future__ import annotations
 
 import json
-import os
-import shutil
-import socket
-import subprocess
 import sys
-import tempfile
-import time
 from pathlib import Path
 
-import httpx
 import pytest
-from PIL import Image
 
-REPO = Path(__file__).resolve().parents[2]
+from tests.e2e._container_harness import (
+    CONTAINER_LIBRARY,
+    Stack,
+    await_job,
+    compose_available,
+    needs_compose,
+    remove_library,
+    temporary_library,
+    write_env_file,
+)
+
 PROJECT = "photo-pipeline-us0803"
 IMAGE = "photo-pipeline-test:us08-03"
 SECRET = "compose-acceptance-secret"
-CONTAINER_LIBRARY = "/library"
-READY_TIMEOUT_SECONDS = 180
-JOB_TIMEOUT_SECONDS = 180
-UP_TIMEOUT_SECONDS = 30 * 60
 
 pytestmark = pytest.mark.container
 
 
-def compose_available() -> bool:
-    try:
-        return (
-            subprocess.run(
-                ["docker", "compose", "version"], capture_output=True, timeout=60
-            ).returncode
-            == 0
-        )
-    except (OSError, subprocess.SubprocessError):
-        return False
-
-
-needs_compose = pytest.mark.skipif(
-    not compose_available(), reason="no Docker daemon with the compose plugin"
-)
-
-
-def free_port() -> int:
-    with socket.socket() as sock:
-        sock.bind(("127.0.0.1", 0))
-        return sock.getsockname()[1]
-
-
-class Stack:
-    """The composition under test, plus the environment it was started with."""
-
-    def __init__(self, library: Path, env_file: Path) -> None:
-        self.library = library
-        self.port = free_port()
-        self.base = f"http://127.0.0.1:{self.port}"
-        # Compose reads the repository's own .env for substitution; the process
-        # environment wins over it, so the test's values are the ones that apply.
-        self.env = {
-            **os.environ,
-            "PHOTO_PIPELINE_IMAGE": IMAGE,
-            "PHOTO_PIPELINE_ENV_FILE": str(env_file),
-            "PHOTO_PIPELINE_LIBRARY_HOST_PATH": str(library),
-            "PHOTO_PIPELINE_LIBRARY_ROOTS": CONTAINER_LIBRARY,
-            "PHOTO_PIPELINE_PORT": str(self.port),
-            "PHOTO_PIPELINE_UID": str(os.getuid()),
-            "PHOTO_PIPELINE_GID": str(os.getgid()),
-        }
-
-    def compose(self, *args: str, check: bool = True, timeout: int = 300):
-        result = subprocess.run(
-            ["docker", "compose", "-p", PROJECT, "-f", str(REPO / "docker-compose.yml"), *args],
-            capture_output=True,
-            text=True,
-            env=self.env,
-            cwd=REPO,
-            timeout=timeout,
-        )
-        if check and result.returncode != 0:
-            raise AssertionError(
-                f"docker compose {' '.join(args)} failed:\n{result.stdout}\n{result.stderr}\n"
-                f"{self.compose('logs', '--tail', '80', check=False).stdout}"
-            )
-        return result
-
-    def wait_until_ready(self) -> None:
-        deadline = time.monotonic() + READY_TIMEOUT_SECONDS
-        while time.monotonic() < deadline:
-            try:
-                if httpx.get(f"{self.base}/api/v1/health/ready", timeout=5).status_code == 200:
-                    return
-            except httpx.HTTPError:
-                pass
-            time.sleep(0.5)
-        logs = self.compose("logs", "--tail", "120", check=False)
-        raise AssertionError(f"the stack never became ready:\n{logs.stdout}\n{logs.stderr}")
-
-    def client(self) -> httpx.Client:
-        client = httpx.Client(base_url=f"{self.base}/api/v1", timeout=60)
-        bootstrap = client.get("/session", headers={"X-Access-Secret": SECRET})
-        assert bootstrap.status_code == 200, bootstrap.text
-        client.headers["X-CSRF-Token"] = bootstrap.json()["csrf_token"]
-        return client
-
-
 @pytest.fixture(scope="module")
 def library() -> Path:
-    """A small fixture library on the host, mounted into both containers.
-
-    Not under pytest's ``tmp_path``: on macOS that is ``/var/folders/...``, which a
-    Docker VM (Colima, Docker Desktop) does not share, so the bind mount would arrive
-    empty and every assertion below would be about nothing. ``$HOME`` is shared by
-    every default configuration.
-    """
-    base = Path(
-        os.environ.get("PHOTO_PIPELINE_TEST_MOUNT_BASE", Path.home() / ".cache" / "photo-pipeline")
-    )
-    base.mkdir(parents=True, exist_ok=True)
-    root = Path(tempfile.mkdtemp(prefix="library-", dir=base))
-    for album, count in (("01_day", 2), ("02_night", 1)):
-        (root / album).mkdir()
-        for index in range(count):
-            colour = (40 * (index + 1), 90, 160)
-            Image.new("RGB", (64, 48), colour).save(root / album / f"{album}_{index}.jpg")
-    # The exclusion sentinel: it must never be discovered, counted, or analyzed.
-    (root / "_IGNORE").mkdir()
-    Image.new("RGB", (32, 32), (0, 0, 0)).save(root / "_IGNORE" / "sentinel.jpg")
+    root = temporary_library({"01_day": 2, "02_night": 1})
     try:
         yield root
     finally:
-        shutil.rmtree(root, ignore_errors=True)
+        remove_library(root)
 
 
 @pytest.fixture(scope="module")
 def env_file(tmp_path_factory) -> Path:
-    """Configuration and secrets come from the environment, so the test writes its
-    own file rather than borrowing the operator's."""
-    path = tmp_path_factory.mktemp("config") / "compose.env"
-    path.write_text(
-        "\n".join(
-            [
-                f"PHOTO_PIPELINE_ACCESS_SECRET={SECRET}",
-                "PHOTO_PIPELINE_LOG_FORMAT=text",
-                # The deterministic vision seam, in the data volume so both roles and
-                # the test can see it (concept §18).
-                "PHOTO_PIPELINE_FAKE_VISION_LOG=/data/vision.log",
-            ]
-        )
-        + "\n"
-    )
-    return path
+    return write_env_file(tmp_path_factory.mktemp("config") / "compose.env", SECRET)
 
 
 @pytest.fixture(scope="module")
 def stack(library, env_file):
     if not compose_available():
         pytest.skip("no Docker daemon with the compose plugin")
-    running = Stack(library, env_file)
-    running.compose("down", "--volumes", "--remove-orphans", check=False)
-    running.compose("up", "--detach", "--build", timeout=UP_TIMEOUT_SECONDS)
+    running = Stack(library, env_file, project=PROJECT, image=IMAGE, secret=SECRET)
+    running.down()
+    running.up("--build")
     try:
         running.wait_until_ready()
         yield running
     finally:
-        running.compose("down", "--volumes", "--remove-orphans", check=False, timeout=300)
-
-
-def await_job(client: httpx.Client, job_id: str, states=("succeeded",)) -> dict:
-    deadline = time.monotonic() + JOB_TIMEOUT_SECONDS
-    snapshot: dict = {}
-    while time.monotonic() < deadline:
-        response = client.get(f"/jobs/{job_id}")
-        if response.status_code == 200:
-            snapshot = response.json()
-            if snapshot["state"] in states:
-                return snapshot
-        time.sleep(0.5)
-    raise AssertionError(f"job {job_id} never reached {states}: {snapshot}")
+        running.down()
 
 
 # ── the mounted library ──────────────────────────────────────────────────────
@@ -296,8 +171,8 @@ def test_the_data_volume_survives_recreating_the_containers(stack):
     finally:
         client.close()
 
-    stack.compose("down", "--remove-orphans", timeout=300)
-    stack.compose("up", "--detach", timeout=UP_TIMEOUT_SECONDS)
+    stack.down(volumes=False)
+    stack.up()
     stack.wait_until_ready()
 
     client = stack.client()

tests/e2e/test_docs_ui.py

@@ -0,0 +1,149 @@
+"""US09-01: the manuals, in the browser, under the application's real policy.
+
+Everything here runs against a real ``photo_pipeline serve`` process serving the
+committed ``docs/`` tree and the vendored renderer — no fixture markdown, no stubbed
+fetch. What that buys is the assertion the story actually cares about: the pages
+render, the links between them work, a diagram becomes a diagram, and the console
+stays empty, including of CSP violations.
+
+The offline contract — reachability, dead links, pinned checksums — is
+``tests/integration/test_documentation.py``.
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+
+import pytest
+
+from tests.e2e._pipeline_harness import Server, seed_library
+
+
+@pytest.fixture(scope="module")
+def server(tmp_path_factory):
+    seeded = seed_library(tmp_path_factory.mktemp("docs"), {"a": 1}, {})
+    running = Server(seeded).start()
+    try:
+        yield running
+    finally:
+        running.stop()
+
+
+@pytest.fixture
+def quiet(page):
+    """Any console error, page error, or failed request fails the test that caused
+    it. A CSP violation arrives as a console error, which is the point."""
+    problems: list[str] = []
+    page.on("console", lambda m: problems.append(m.text) if m.type == "error" else None)
+    page.on("pageerror", lambda error: problems.append(str(error)))
+    page.on("requestfailed", lambda request: problems.append(f"failed request: {request.url}"))
+    yield problems
+
+
+def test_the_documentation_opens_from_the_navigation(page, server, quiet):
+    page.goto(f"{server.base}/app/#/workflow")
+    page.locator('nav a[data-nav="docs"]').click()
+
+    page.get_by_test_id("doc").wait_for()
+    assert page.get_by_test_id("doc").get_attribute("data-page") == "index"
+    # The sidebar is the index's own reading order, not a second list to maintain.
+    assert page.get_by_test_id("doc-pages").get_by_role("link", name="Overview").is_visible()
+    assert quiet == []
+
+
+def test_a_link_between_documents_stays_inside_the_application(page, server, quiet):
+    page.goto(f"{server.base}/app/#/docs")
+    page.get_by_test_id("doc").wait_for()
+    page.get_by_test_id("doc").get_by_role("link", name="Overview").click()
+
+    page.wait_for_selector('[data-testid="doc"][data-page="overview"]')
+    assert "#/docs?page=overview" in page.url, "a relative .md link must not leave the app"
+    # And back again, by the link the document itself carries.
+    page.get_by_test_id("doc").get_by_role("link", name="← Documentation index").click()
+    page.wait_for_selector('[data-testid="doc"][data-page="index"]')
+    assert quiet == []
+
+
+def test_a_deep_link_to_a_heading_lands_on_that_heading(page, server, quiet):
+    page.goto(f"{server.base}/app/#/docs?page=overview&anchor=the-workflow")
+    heading = page.locator("#the-workflow")
+    heading.wait_for()
+
+    assert heading.inner_text().strip() == "The workflow"
+    assert heading.evaluate("node => node.getBoundingClientRect().top < window.innerHeight")
+    assert quiet == []
+
+
+def test_a_mermaid_block_becomes_a_diagram_under_the_unchanged_script_policy(page, server, quiet):
+    page.goto(f"{server.base}/app/#/docs?page=overview")
+    diagram = page.get_by_test_id("diagram").first
+    diagram.wait_for()
+
+    # A real drawing, not the source text and not an error node.
+    assert diagram.locator("svg").count() == 1
+    assert diagram.evaluate("node => node.querySelector('svg').getBBox().width") > 100
+    assert page.locator("pre code.language-mermaid").count() == 0
+    assert quiet == [], "rendering a diagram must not violate the policy"
+
+    policy = page.evaluate(
+        "async () => (await fetch('/app/')).headers.get('content-security-policy')"
+    )
+    assert "script-src 'self';" in policy
+    assert "unsafe-eval" not in policy
+
+
+def test_every_diagram_in_the_architecture_overview_draws(page, server, quiet):
+    """Five diagrams, three of them state machines (US09-03). A mermaid block with a
+    syntax error renders an error node instead of throwing, so a page that merely
+    loaded proves nothing — count the drawings and read the source for the states."""
+    source = (Path(__file__).resolve().parents[2] / "docs" / "architecture.md").read_text()
+    expected = source.count("```mermaid")
+    assert expected >= 5, "the overview lost its diagrams"
+
+    page.goto(f"{server.base}/app/#/docs?page=architecture")
+    page.get_by_test_id("diagram").first.wait_for()
+    page.wait_for_function(
+        "count => document.querySelectorAll('[data-testid=\"diagram\"] svg').length === count",
+        arg=expected,
+    )
+
+    drawn = page.locator('[data-testid="diagram"] svg')
+    assert drawn.count() == expected
+    for index in range(expected):
+        diagram = drawn.nth(index)
+        # Labels may be <text> or HTML inside a <foreignObject> depending on the
+        # diagram type, so ask for the rendered text rather than a specific element.
+        labels = diagram.evaluate("node => node.textContent.trim().length")
+        assert labels > 0, f"diagram {index} drew no labels"
+        assert diagram.locator(".error-icon, .error-text").count() == 0, f"diagram {index} errored"
+    assert page.locator("pre code.language-mermaid").count() == 0
+    assert quiet == []
+
+
+def test_an_unknown_page_says_so_without_naming_a_path(page, server, quiet):
+    page.goto(f"{server.base}/app/#/docs?page=no-such-manual")
+    message = page.get_by_test_id("doc-not-found")
+    message.wait_for()
+
+    text = message.inner_text()
+    assert "does not exist" in text
+    assert "/" not in text.replace("Back to the documentation index", ""), text
+    message.get_by_role("link").click()
+    page.wait_for_selector('[data-testid="doc"][data-page="index"]')
+    # The missing page is a 404 and the browser says so; nothing else may go wrong,
+    # and in particular the view must not throw on the way to its own message.
+    assert all("404" in problem for problem in quiet), quiet
+
+
+def test_the_documentation_is_readable_without_a_session(page, server, quiet):
+    """The troubleshooting page is needed most by whoever is locked out."""
+    page.goto(f"{server.base}/app/#/docs")
+    page.get_by_test_id("doc").wait_for()
+    unauthenticated = page.evaluate(
+        """async () => {
+             const response = await fetch('/docs/index.md', { credentials: 'omit' });
+             return { status: response.status, length: (await response.text()).length };
+           }"""
+    )
+    assert unauthenticated["status"] == 200
+    assert unauthenticated["length"] > 100

tests/e2e/test_phase_h_container.py

@@ -0,0 +1,510 @@
+"""Phase H — the container acceptance gate (US08-05).
+
+The deployed application is verified the way the host application is: real image,
+real composition, real browser, real restarts. Everything below runs against
+containers that this suite provisions from the built image, against a temporary
+fixture library on a bind mount and an isolated data volume, and destroys afterwards.
+
+Four journeys, one per thing a deployment can get wrong:
+
+* **the browser journey** — discovery, duplicate review, analysis, album proposal,
+  rename, upload preflight, and archive, driven through the containerized frontend;
+* **the upgrade journey** — the previous version's image runs first, then this one,
+  and the database, its migrations, the rename journal, the job history, and the
+  thumbnail cache have to still be there;
+* **the restart journey** — both containers are killed mid-job and the work resumes
+  without doing anything twice;
+* **the security gates** — the refusals a loopback deployment made are still made
+  behind a published port: no session, forged forwarded headers, a path that leaves
+  the mounted library, and a secret in the logs.
+
+Run it as one command, with evidence: ``python -m photo_pipeline container-gate``.
+"""
+
+from __future__ import annotations
+
+import collections
+from contextlib import closing
+import json
+import os
+import shutil
+import subprocess
+import sys
+import time
+from pathlib import Path
+
+import httpx
+import pytest
+from playwright.sync_api import expect
+
+from photo_pipeline.db import head_revision
+from tests.e2e._container_harness import (
+    Stack,
+    await_job,
+    build_image,
+    compose_available,
+    needs_compose,
+    remove_library,
+    temporary_library,
+    write_env_file,
+)
+
+pytestmark = [pytest.mark.phase_h, pytest.mark.container]
+
+PROJECT = "photo-pipeline-us0805"
+IMAGE = "photo-pipeline-test:us08-05"
+# The version being upgraded *from*. CI passes the tag it last published; without one,
+# the previous commit's tree is built, which is the same claim without a registry.
+PREVIOUS_IMAGE = os.environ.get("PHOTO_PIPELINE_PREVIOUS_IMAGE")
+PREVIOUS_TAG = "photo-pipeline-test:us08-05-previous"
+SECRET = "container-gate-access-secret"
+IMMICH_SENTINEL = "immich-sentinel-9f3a2b"
+HOSTNAME = "photos.test"
+ALBUM = "rome"
+RENAMED = "2019 Rome"
+BURST = 30
+
+
+# ── the image and the stacks ─────────────────────────────────────────────────
+
+
+@pytest.fixture(scope="session")
+def image() -> str:
+    if not compose_available():
+        pytest.skip("no Docker daemon with the compose plugin")
+    return build_image(IMAGE)
+
+
+@pytest.fixture(scope="module")
+def journey(image, tmp_path_factory):
+    """One album, one exact duplicate of a photo in it, and the exclusion sentinel."""
+    library = temporary_library({ALBUM: 3}, prefix="us0805-journey-")
+    shutil.copyfile(library / ALBUM / f"{ALBUM}_0.jpg", library / ALBUM / "copy.jpg")
+    env_file = write_env_file(
+        tmp_path_factory.mktemp("journey") / "gate.env",
+        SECRET,
+        {
+            "PHOTO_PIPELINE_ALLOWED_HOSTS": HOSTNAME,
+            # Configured but never reachable: the upload view's job here is to show the
+            # preflight blockers, and the key's job is to be absent from every log.
+            "PHOTO_PIPELINE_IMMICH_API_KEY": IMMICH_SENTINEL,
+            "PHOTO_PIPELINE_IMMICH_SERVER_URL": "http://127.0.0.1:1",
+        },
+    )
+    stack = Stack(library, env_file, project=f"{PROJECT}-journey", image=image, secret=SECRET)
+    try:
+        stack.down()
+        stack.up()
+        stack.wait_until_ready()
+        _seed_journey(stack)
+        yield stack
+    finally:
+        stack.down()
+        remove_library(library)
+
+
+def _seed_journey(stack: Stack) -> None:
+    """Everything the views need, established over the public API before they render."""
+    with closing(stack.client()) as client:
+        client.post("/inventory/scan").raise_for_status()
+        client.post("/duplicates/detect").raise_for_status()
+        queue = client.get("/safety/queue", params={"limit": 100}).json()["items"]
+        for item in queue:
+            decided = client.post(
+                "/safety/decisions", json={"asset_id": item["asset_id"], "decision": "sfw"}
+            )
+            assert decided.status_code == 200, decided.text
+        job = client.post("/analysis/jobs").json()
+        await_job(client, job["id"])
+        client.post("/albums/proposals", json={}).raise_for_status()
+        # An archive destination inside the data volume: a second mount would prove
+        # nothing more, and the view needs a location to have something to show.
+        stack.compose("exec", "-T", "api", "mkdir", "-p", "/data/archive")
+        client.post(
+            "/archive-locations", json={"name": "external", "root": "/data/archive"}
+        ).raise_for_status()
+
+
+# ── the browser journey ──────────────────────────────────────────────────────
+
+
+@needs_compose
+def test_the_browser_journey_covers_every_stage_view_of_the_deployed_app(page, journey):
+    """One pass through the deployed frontend, in workflow order.
+
+    The access secret is supplied the way a person supplies it — the app asks, the
+    answer is kept for the tab — so what is proven is the authenticated deployment,
+    not a test-only bypass.
+    """
+    page.on("dialog", lambda dialog: dialog.accept(SECRET))
+    errors: list[str] = []
+    page.on("pageerror", lambda error: errors.append(str(error)))
+    base = journey.base
+    with closing(journey.client()) as client:
+        cluster = client.get("/duplicates/clusters").json()["items"][0]
+
+    # ── discovery ────────────────────────────────────────────────────────────
+    page.goto(f"{base}/app/#/workflow")
+    page.get_by_test_id("stage-safety").wait_for()
+    expect(page.get_by_test_id("stage-analysis")).to_be_visible()
+
+    page.goto(f"{base}/app/#/inventory")
+    rows = page.get_by_test_id("asset-row")
+    rows.first.wait_for()
+    assert rows.count() == 4, "three photos and the duplicate copy; never the sentinel"
+    assert "sentinel" not in page.content() and "_IGNORE" not in page.content()
+
+    # ── duplicate review ─────────────────────────────────────────────────────
+    page.goto(f"{base}/app/#/duplicates/{cluster['id']}")
+    page.get_by_test_id("cluster-state").wait_for()
+    # Exact bytes: the cluster arrives decided, and the review surface has to show
+    # both members and the evidence the decision was made on.
+    expect(page.get_by_test_id("cluster-state")).to_contain_text("decided")
+    expect(page.get_by_test_id("member")).to_have_count(2)
+    expect(page.get_by_test_id("member").first).to_contain_text("/library/")
+
+    # ── analysis ─────────────────────────────────────────────────────────────
+    page.goto(f"{base}/app/#/analyze")
+    page.get_by_test_id("analyze-counts").wait_for()
+    expect(page.get_by_test_id("run-analysis")).to_be_visible()
+
+    # ── album proposal ───────────────────────────────────────────────────────
+    page.goto(f"{base}/app/#/albums?album={ALBUM}")
+    page.get_by_test_id("suggested-name").wait_for()
+    page.get_by_test_id("final-name").fill(RENAMED)
+    page.get_by_test_id("save-name").click()
+    page.get_by_test_id("approve").click()
+    expect(page.get_by_test_id("proposal-status")).to_contain_text("approved")
+
+    # ── rename, applied against the bind mount ───────────────────────────────
+    page.goto(f"{base}/app/#/renames")
+    page.get_by_test_id("build-plan").click()
+    page.get_by_test_id("operations").wait_for()
+    expect(page.get_by_test_id("op-destination").first).to_contain_text(RENAMED)
+    page.get_by_test_id("apply-plan").click()
+    expect(page.get_by_test_id("apply-result")).to_contain_text("Applied 1, failed 0")
+    # The mounted library is the host's directory: the container renamed the operator's
+    # folder, not a copy inside its own layer.
+    assert (journey.library / RENAMED).is_dir()
+    assert not (journey.library / ALBUM).exists()
+
+    # ── upload preflight ─────────────────────────────────────────────────────
+    page.goto(f"{base}/app/#/uploads")
+    page.get_by_test_id("upload-scope").wait_for()
+    expect(page.get_by_test_id("album-row").first).to_be_visible()
+    assert IMMICH_SENTINEL not in page.content(), "the API key never reaches the browser"
+
+    # ── archive ──────────────────────────────────────────────────────────────
+    page.goto(f"{base}/app/#/archive")
+    page.get_by_test_id("archive-locations").wait_for()
+    expect(page.get_by_test_id("location-row").first).to_contain_text("external")
+
+    assert errors == [], f"the deployed frontend raised page errors: {errors}"
+
+
+# ── the upgrade journey ──────────────────────────────────────────────────────
+
+
+@pytest.fixture(scope="module")
+def previous_image(image) -> str:
+    """The image the deployment is upgrading *from*.
+
+    The published tag when there is one. Before the first publish there is nothing to
+    pull, and refusing then would mean the gate could never let the first deploy
+    through — so the previous commit's tree is built instead, which is the same claim
+    without a registry.
+    """
+    if PREVIOUS_IMAGE:
+        pulled = subprocess.run(
+            ["docker", "pull", PREVIOUS_IMAGE], capture_output=True, timeout=1800
+        )
+        if pulled.returncode == 0:
+            return PREVIOUS_IMAGE
+    return build_image(PREVIOUS_TAG, revision="HEAD~1")
+
+
+@needs_compose
+def test_an_upgrade_keeps_the_database_the_journal_the_jobs_and_the_cache(
+    image, previous_image, tmp_path_factory
+):
+    library = temporary_library({ALBUM: 2}, prefix="us0805-upgrade-")
+    env_file = write_env_file(tmp_path_factory.mktemp("upgrade") / "gate.env", SECRET)
+    stack = Stack(
+        library, env_file, project=f"{PROJECT}-upgrade", image=previous_image, secret=SECRET
+    )
+    try:
+        stack.down()
+        stack.up()
+        stack.wait_until_ready()
+
+        # ── what the previous version leaves behind ──────────────────────────
+        with closing(stack.client()) as client:
+            client.post("/inventory/scan").raise_for_status()
+            assets = client.get("/inventory/assets", params={"limit": 200}).json()["items"]
+            for item in client.get("/safety/queue", params={"limit": 100}).json()["items"]:
+                client.post(
+                    "/safety/decisions", json={"asset_id": item["asset_id"], "decision": "sfw"}
+                ).raise_for_status()
+            job = client.post("/analysis/jobs").json()
+            finished = await_job(client, job["id"])
+            for asset in assets:  # populate the thumbnail cache
+                thumbnail = client.get(f"/assets/{asset['id']}/thumbnail", params={"size": 256})
+                assert thumbnail.status_code == 200, thumbnail.text
+            # A rename plan, left unapplied: the journal has to survive the upgrade
+            # exactly as it was, or a half-applied one could not be recovered.
+            client.post("/albums/proposals", json={}).raise_for_status()
+            _approve(client, RENAMED)
+            plan = client.post("/rename-plans").json()
+
+        before = {
+            "assets": sorted(asset["id"] for asset in assets),
+            "job": finished["progress"],
+            "plan": (plan["id"], plan["checksum"]),
+            "thumbnails": _cache_files(stack),
+            "revision": _revision(stack),
+        }
+        assert before["thumbnails"], "no thumbnail was cached, so nothing would be proven"
+
+        # ── the upgrade: same volume, new image ──────────────────────────────
+        stack.down(volumes=False)
+        stack.use_image(image)
+        stack.up()
+        stack.wait_until_ready()
+
+        after_revision = _revision(stack)
+        assert after_revision == head_revision(), "the new image did not migrate the volume"
+        if after_revision != before["revision"]:
+            # A schema change is snapshotted before it is applied (US07-05), so a
+            # failed upgrade is restorable rather than a lost library.
+            assert _ls(stack, "/data/backups"), "a migration ran without a backup"
+
+        with closing(stack.client()) as client:
+            assets = client.get("/inventory/assets", params={"limit": 200}).json()["items"]
+            assert sorted(asset["id"] for asset in assets) == before["assets"]
+            assert client.get(f"/jobs/{job['id']}").json()["progress"] == before["job"]
+            plans = client.get("/rename-plans").json()["items"]
+            assert (plans[0]["id"], plans[0]["checksum"]) == before["plan"]
+            for asset in assets:
+                assert (
+                    client.get(f"/analysis/results/{asset['id']}").status_code == 200
+                ), "an analysis result did not survive the upgrade"
+        # The cache is keyed by pixel hash and thumbnail version, so an upgrade that
+        # kept the volume must keep the files: regenerating them is work nobody asked
+        # for, and losing them silently is how a cache stops being one.
+        assert set(before["thumbnails"]) <= set(_cache_files(stack))
+    finally:
+        stack.down()
+        remove_library(library)
+
+
+def _approve(client: httpx.Client, name: str, *, album: str = ALBUM) -> None:
+    for payload, route in (({"name": name}, "edit"), ({}, "approve")):
+        current = client.get(f"/albums/proposals/{album}").json()
+        client.post(
+            f"/albums/proposals/{album}/{route}",
+            json={**payload, "expected_version": current["version"]},
+        ).raise_for_status()
+
+
+def _exec(stack: Stack, *args: str) -> str:
+    return stack.compose("exec", "-T", "api", *args).stdout
+
+
+def _ls(stack: Stack, directory: str) -> list[str]:
+    listing = stack.compose("exec", "-T", "api", "ls", directory, check=False)
+    return [line for line in listing.stdout.split() if line]
+
+
+def _cache_files(stack: Stack) -> list[str]:
+    return sorted(
+        _exec(stack, "find", "/data/cache", "-type", "f", "-name", "*.webp").split()
+    )
+
+
+def _revision(stack: Stack) -> str:
+    """The schema revision the volume's database is actually at."""
+    return _exec(
+        stack,
+        "python",
+        "-c",
+        "import sqlite3;print(sqlite3.connect('/data/photo_pipeline.db')"
+        ".execute('select version_num from alembic_version').fetchone()[0])",
+    ).strip()
+
+
+# ── the restart journey ──────────────────────────────────────────────────────
+
+
+@needs_compose
+def test_killing_both_containers_mid_job_resumes_without_doing_anything_twice(
+    image, tmp_path_factory
+):
+    """`docker kill` is the honest restart: no grace period, no orderly stop, no
+    chance for either process to write a tidy final state."""
+    library = temporary_library({"burst": BURST}, prefix="us0805-restart-")
+    env_file = write_env_file(tmp_path_factory.mktemp("restart") / "gate.env", SECRET)
+    stack = Stack(library, env_file, project=f"{PROJECT}-restart", image=image, secret=SECRET)
+    try:
+        stack.down()
+        stack.up()
+        stack.wait_until_ready()
+
+        with closing(stack.client()) as client:
+            client.post("/inventory/scan").raise_for_status()
+            assets = client.get("/inventory/assets", params={"limit": 200}).json()["items"]
+            assert len(assets) == BURST
+            for item in client.get("/safety/queue", params={"limit": 100}).json()["items"]:
+                client.post(
+                    "/safety/decisions", json={"asset_id": item["asset_id"], "decision": "sfw"}
+                ).raise_for_status()
+            job = client.post("/analysis/jobs").json()
+            progress = _wait_for_progress(client, job["id"])
+
+        assert 0 < progress["done"] < progress["total"], progress
+        stack.compose("kill", "api", "worker")
+
+        stack.up()
+        stack.wait_until_ready()
+        with closing(stack.client()) as client:
+            finished = await_job(client, job["id"])
+            assert finished["progress"]["done"] == BURST, finished
+            results = [
+                client.get(f"/analysis/results/{asset['id']}").json() for asset in assets
+            ]
+
+        # Exactly one stored result per asset: at-least-once execution, idempotent
+        # recovery — a retried item overwrites its own attempt, it does not add one.
+        assert len(results) == BURST
+        assert all(result["description"] for result in results)
+        # And the side effect nobody can take back — the call to the provider — happened
+        # again only for whatever was in flight when the containers died.
+        analysed = collections.Counter(_exec(stack, "cat", "/data/vision.log").split())
+        assert len(analysed) == BURST, "every photo was analysed, and only the library's"
+        assert max(analysed.values()) <= 2, dict(analysed)
+        assert sum(1 for count in analysed.values() if count > 1) <= 1, dict(analysed)
+    finally:
+        stack.down()
+        remove_library(library)
+
+
+def _wait_for_progress(client: httpx.Client, job_id: str, *, timeout: float = 120) -> dict:
+    """Wait until the job is provably under way but provably unfinished."""
+    deadline = time.monotonic() + timeout
+    while time.monotonic() < deadline:
+        snapshot = client.get(f"/jobs/{job_id}").json()
+        progress = snapshot["progress"]
+        if progress["done"] and progress["done"] < progress["total"]:
+            return progress
+        if snapshot["state"] in ("succeeded", "failed"):
+            raise AssertionError(f"the job finished before it could be interrupted: {snapshot}")
+        time.sleep(0.05)
+    raise AssertionError(f"the job never started: {client.get(f'/jobs/{job_id}').json()}")
+
+
+# ── the security gates ───────────────────────────────────────────────────────
+
+
+@needs_compose
+def test_the_deployed_instance_refuses_a_caller_without_a_session(journey):
+    with httpx.Client(base_url=f"{journey.base}/api/v1", timeout=30) as client:
+        for method, path in (("GET", "/workflow"), ("POST", "/inventory/scan")):
+            response = client.request(method, path)
+            assert response.status_code == 401, path
+            assert response.json()["error"]["code"] == "unauthenticated"
+        # A session still has to be paid for with the operator's secret.
+        assert client.get("/session", headers={"X-Access-Secret": "guessed"}).status_code == 401
+        # Readiness stays open: the orchestrator's health check holds no session.
+        assert client.get("/health/ready").status_code == 200
+
+
+@needs_compose
+def test_forged_forwarded_headers_cannot_smuggle_an_allowed_host_past_the_check(journey):
+    """Behind a proxy the app believes ``X-Forwarded-*`` — but only from the proxy.
+    Nothing in this composition is a trusted proxy, so the claim is the client's."""
+    with httpx.Client(base_url=f"{journey.base}/api/v1", timeout=30) as client:
+        client.headers["X-CSRF-Token"] = (
+            client.get("/session", headers={"X-Access-Secret": SECRET}).json()["csrf_token"]
+        )
+        # The hostname this deployment is reached under is accepted.
+        assert client.get("/workflow", headers={"Host": HOSTNAME}).status_code == 200
+
+        forged = client.get(
+            "/workflow",
+            headers={"Host": "photos.evil.example", "X-Forwarded-Host": HOSTNAME},
+        )
+        assert forged.status_code == 403
+        assert forged.json()["error"]["code"] == "host_not_allowed"
+        # A forged protocol claim must not mark the session cookie as HTTPS-only
+        # either — that would strand the operator's real, plain-HTTP session.
+        bootstrap = httpx.get(
+            f"{journey.base}/api/v1/session",
+            headers={"X-Access-Secret": SECRET, "X-Forwarded-Proto": "https"},
+            timeout=30,
+        )
+        assert "secure" not in bootstrap.headers["set-cookie"].lower()
+
+
+@needs_compose
+def test_a_path_that_leaves_the_mounted_library_is_refused(journey):
+    """The container's own filesystem is not the library. A symlink swapped under a
+    known asset is the sharpest version of the question, because the database still
+    points at a path inside the mount."""
+    with closing(journey.client()) as client:
+        asset = client.get("/inventory/assets", params={"limit": 200}).json()["items"][0]
+        original = journey.library / Path(asset["current_path"]).relative_to("/library")
+        kept = original.read_bytes()
+        original.unlink()
+        original.symlink_to("/etc/passwd")
+        try:
+            escaped = client.get(f"/assets/{asset['id']}/thumbnail", params={"size": 256})
+        finally:
+            original.unlink()
+            original.write_bytes(kept)
+
+        assert escaped.status_code == 403
+        assert escaped.json()["error"]["code"] == "path_not_allowed"
+        assert "root:" not in escaped.text
+
+        # And a root that names nothing mounted is refused before the process serves.
+        refused = journey.compose(
+            "run",
+            "--rm",
+            "--no-deps",
+            "--env",
+            "PHOTO_PIPELINE_LIBRARY_ROOTS=/srv/photos",
+            "api",
+            "serve",
+            check=False,
+        )
+        assert refused.returncode == 5, refused.stdout + refused.stderr
+
+
+@needs_compose
+def test_no_secret_reaches_the_container_logs(journey):
+    with httpx.Client(base_url=f"{journey.base}/api/v1", timeout=30) as client:
+        client.get("/session", headers={"X-Access-Secret": "wrong-secret-attempt"})
+        client.get("/session", headers={"X-Access-Secret": SECRET})
+    logs = journey.logs()
+    assert SECRET not in logs
+    assert IMMICH_SENTINEL not in logs
+    assert "wrong-secret-attempt" not in logs, "a rejected secret is still a secret"
+    # The refusal itself is logged, so an operator can see the attempt.
+    assert "access secret rejected" in logs
+
+
+@needs_compose
+def test_the_evidence_of_this_run_names_the_stack_it_was_produced_from(journey):
+    """A gate that cannot say what it ran against is an opinion. `docker compose ps`
+    is the record: one API, one worker, one completed migration."""
+    listing = [
+        json.loads(line)
+        for line in journey.compose("ps", "--all", "--format", "json").stdout.splitlines()
+        if line.strip()
+    ]
+    services = {entry["Service"]: entry["State"] for entry in listing}
+    assert services["api"] == "running" and services["worker"] == "running"
+    assert services["migrate"] == "exited"
+
+
+if __name__ == "__main__":  # a quick way to run just this file
+    raise SystemExit(pytest.main([__file__, "-v", *sys.argv[1:]]))

tests/integration/test_architecture_overview.py

@@ -0,0 +1,164 @@
+"""US09-03: the architecture overview, checked against the architecture.
+
+An architecture document is the one that rots most quietly: nothing breaks when it
+describes a module that was renamed two epics ago, it just quietly misleads the next
+person. So the parts of it that the code also knows — module names, state machines,
+table names, the modules an invariant is claimed to live in — are compared with the
+code, and a missing one fails the suite.
+
+The diagrams' rendering is proven in ``tests/e2e/test_docs_ui.py``.
+"""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+
+import photo_pipeline.models  # noqa: F401  (registers every table on Base.metadata)
+from photo_pipeline.db import Base
+from photo_pipeline.services import jobs, rename_journal, upload_batches
+
+REPO = Path(__file__).resolve().parents[2]
+PACKAGE = REPO / "photo_pipeline"
+OVERVIEW = REPO / "docs" / "architecture.md"
+TEXT = OVERVIEW.read_text()
+
+# Names the map does not owe the reader individually: the package markers, the CLI
+# entry point, and the two modules that exist to be small and obvious.
+UNMAPPED = {"__init__", "__main__", "logging"}
+
+
+def modules() -> set[str]:
+    """Every module and package under photo_pipeline, by the name it is imported as."""
+    names = set()
+    for path in PACKAGE.rglob("*.py"):
+        relative = path.relative_to(PACKAGE)
+        names.add(relative.parts[0] if len(relative.parts) > 1 else relative.stem)
+    return {name.removesuffix(".py") for name in names} - UNMAPPED
+
+
+def states(container: type) -> set[str]:
+    return {
+        value
+        for name, value in vars(container).items()
+        if name.isupper() and isinstance(value, str)
+    }
+
+
+# ── the module map ───────────────────────────────────────────────────────────
+
+
+def test_every_module_appears_in_the_map():
+    """A service nobody documented is a service the next person re-implements."""
+    missing = sorted(name for name in modules() if name not in TEXT)
+    assert missing == [], f"absent from the architecture overview: {missing}"
+
+
+def test_every_service_is_named_individually():
+    """The package table says what `services/` is for; this is the list that actually
+    drifts, because a new service is added roughly every story."""
+    services = {path.stem for path in (PACKAGE / "services").glob("*.py")} - UNMAPPED
+    missing = sorted(name for name in services if not re.search(rf"\b{name}\b", TEXT))
+    assert missing == [], f"services missing from the overview: {missing}"
+
+
+def test_the_map_names_no_module_that_does_not_exist():
+    """Every `services/<name>` or backticked module path in the document resolves."""
+    referenced = set(re.findall(r"`(?:photo_pipeline/)?([a-z_]+)/([a-z_]+)\.py`", TEXT))
+    referenced |= {("", name) for name in re.findall(r"`([a-z_]+)\.py`", TEXT)}
+    missing = [
+        f"{package}/{module}.py" if package else f"{module}.py"
+        for package, module in referenced
+        if not (PACKAGE / package / f"{module}.py").is_file()
+    ]
+    assert missing == [], f"documented but absent from the source: {missing}"
+
+
+# ── the state machines ───────────────────────────────────────────────────────
+
+
+def test_every_job_state_is_documented():
+    missing = sorted(state for state in states(jobs.JobState) if state not in TEXT)
+    assert missing == [], f"job states missing from the overview: {missing}"
+
+
+def test_every_rename_journal_state_is_documented():
+    missing = sorted(state for state in states(rename_journal.JournalState) if state not in TEXT)
+    assert missing == [], f"journal states missing from the overview: {missing}"
+
+
+def test_every_upload_batch_state_is_documented():
+    missing = sorted(state for state in states(upload_batches.BatchState) if state not in TEXT)
+    assert missing == [], f"batch states missing from the overview: {missing}"
+
+
+def test_the_unsafe_journal_states_are_named_as_the_ones_that_block():
+    """The reason an unrelated mutation is refused has to be findable."""
+    for state in rename_journal.UNSAFE_STATES:
+        assert state in TEXT
+    assert "rename_recovery_required" in TEXT
+
+
+def test_an_uncertain_upload_is_documented_as_not_retryable():
+    assert upload_batches.BatchState.UNKNOWN not in upload_batches.RUNNABLE_STATES
+    assert "not** restartable" in TEXT or "not restartable" in TEXT
+
+
+# ── the data model ───────────────────────────────────────────────────────────
+
+
+def test_every_table_is_listed():
+    missing = sorted(name for name in Base.metadata.tables if name not in TEXT)
+    assert missing == [], f"tables missing from the overview: {missing}"
+
+
+def test_the_document_lists_no_table_that_does_not_exist():
+    listed = set(re.findall(r"`([a-z_]+)`", TEXT))
+    plausible = {name for name in listed if name.endswith("s") and "_" in name}
+    invented = sorted(
+        name
+        for name in plausible
+        if name not in Base.metadata.tables
+        and not (PACKAGE / "services" / f"{name}.py").is_file()
+        and name not in {"library_roots", "trusted_proxies", "allowed_hosts", "asset_paths"}
+    )
+    assert invented == [], f"looks like a table but is not one: {invented}"
+
+
+# ── the invariants ───────────────────────────────────────────────────────────
+
+
+def test_each_invariant_points_at_a_module_that_enforces_it():
+    """The point of the table is to answer 'where do I look'. A wrong answer there
+    costs more than no answer."""
+    claims = {
+        "path_policy.is_excluded": PACKAGE / "path_policy.py",
+        "path_policy.resolve_in_roots": PACKAGE / "path_policy.py",
+        "services/app_lock.py": PACKAGE / "services" / "app_lock.py",
+        "services/exif_checkpoint.py": PACKAGE / "services" / "exif_checkpoint.py",
+        "services/archive_transfer.py": PACKAGE / "services" / "archive_transfer.py",
+        "api/security.py": PACKAGE / "api" / "security.py",
+        "imaging.py": PACKAGE / "imaging.py",
+    }
+    for claim, path in claims.items():
+        assert claim in TEXT, f"the invariant table does not mention {claim}"
+        assert path.is_file(), f"{claim} does not exist"
+        function = claim.rpartition(".")[2] if "/" not in claim else ""
+        if function and function != "py":
+            assert f"def {function}" in path.read_text(), f"{claim} is not defined there"
+
+
+def test_the_lock_order_matches_the_ranks_in_the_code():
+    from photo_pipeline.jobs import locks
+
+    order = [name for name, _ in sorted(locks.LOCK_RANK.items(), key=lambda item: item[1])]
+    assert order[0].startswith("library"), "the broadest lock is no longer the library"
+    assert "library → stage/job → album/folder → asset" in TEXT
+
+
+def test_the_frozen_donor_archive_is_described_as_provenance_not_a_dependency():
+    """That production never imports the frozen sources is proven by
+    ``tests/unit/test_legacy_archive.py`` — which also forbids any other test from
+    naming that directory, so this one checks the claim by its consequence."""
+    assert "must not import" in TEXT
+    assert "provenance and rollback evidence" in TEXT

tests/integration/test_container_gate.py

@@ -0,0 +1,147 @@
+"""US08-05: the container acceptance gate's own contract, checked without a daemon.
+
+The gate itself needs Docker, a browser, and several minutes; running it from inside
+the suite would be a fork bomb with better manners. What is checkable offline is what
+makes it a *gate* rather than a long test run:
+
+* it selects the container journeys by marker, so adding one is enough to put it in
+  front of a deploy;
+* it accepts no skip at all — every reason a check would skip here (no daemon, no
+  compose plugin, no browser) means the deployed runtime was not proven;
+* it retains checksummed evidence per run;
+* and CI runs it on `main`, which is what the publish step waits for.
+
+The running proof is ``tests/e2e/test_phase_h_container.py``.
+"""
+
+from __future__ import annotations
+
+import json
+import re
+from pathlib import Path
+
+import yaml
+
+from photo_pipeline.config import Config
+from photo_pipeline.services import release
+
+REPO = Path(__file__).resolve().parents[2]
+SUITE = REPO / "tests" / "e2e" / "test_phase_h_container.py"
+TEST_WORKFLOW = yaml.safe_load((REPO / ".gitea" / "workflows" / "test.yml").read_text())
+
+
+def _config(tmp_path) -> Config:
+    for name in ("data", "lib"):
+        (tmp_path / name).mkdir(exist_ok=True)
+    return Config.from_env(
+        {
+            "PHOTO_PIPELINE_DATA_DIR": str(tmp_path / "data"),
+            "PHOTO_PIPELINE_LIBRARY_ROOTS": str(tmp_path / "lib"),
+        }
+    )
+
+
+# ── what the gate runs ───────────────────────────────────────────────────────
+
+
+def test_the_gate_selects_the_container_journeys_by_their_marker():
+    assert release.CONTAINER_STAGES == (("container", ("tests/e2e", "-m", "phase_h")),)
+    declared = [line for line in (REPO / "pyproject.toml").read_text().splitlines()
+                if line.strip().startswith('"phase_h:')]
+    assert declared, "an unregistered marker selects nothing and fails no gate"
+
+
+def test_every_journey_in_the_suite_carries_the_marker():
+    """A test in that file without the marker is a check the gate never runs."""
+    source = SUITE.read_text()
+    assert "pytestmark = [pytest.mark.phase_h, pytest.mark.container]" in source
+    journeys = re.findall(r"^def (test_[a-z_]+)", source, re.MULTILINE)
+    assert len(journeys) >= 4, journeys
+    for required in ("browser", "upgrade", "kill", "secret"):
+        assert any(required in name for name in journeys), (required, journeys)
+
+
+# ── no skip is an environment limit here ─────────────────────────────────────
+
+
+def test_the_gate_accepts_no_skipped_check_at_all():
+    assert release.CONTAINER_ALLOWED_SKIP_REASONS == ()
+    docker_missing = release.StageResult(
+        "container", [], 0, 0.1, "1 skipped", ["SKIPPED [1] x.py:1: no Docker daemon available"]
+    )
+    # The release gate tolerates exactly this one, because the image definition is
+    # still checked offline. The container gate cannot: it is the deployment's proof.
+    assert release.unexpected_skips([docker_missing]) == []
+    assert release.unexpected_skips(
+        [docker_missing], release.CONTAINER_ALLOWED_SKIP_REASONS
+    ) == ["SKIPPED [1] x.py:1: no Docker daemon available"]
+
+
+def test_a_skipped_check_fails_the_run_and_the_evidence_says_so(tmp_path):
+    skipping = tmp_path / "test_skipping.py"
+    skipping.write_text(
+        "import pytest\n\n"
+        "def test_x():\n"
+        "    pytest.skip('no Docker daemon available')\n"
+    )
+    evidence = tmp_path / "evidence"
+
+    report = release.run_gate(
+        _config(tmp_path),
+        output=evidence,
+        stages=(("container", (str(skipping),)),),
+        allowed_skips=release.CONTAINER_ALLOWED_SKIP_REASONS,
+    )
+
+    assert report["ok"] is False
+    assert report["failures"] == [], "the stage passed; the skip is what fails the gate"
+    assert report["unexpected_skips"], report
+    written = json.loads((evidence / release.REPORT_NAME).read_text())
+    assert written["ok"] is False
+    assert (evidence / "logs" / "container.log").exists()
+    for line in (evidence / release.CHECKSUMS_NAME).read_text().splitlines():
+        digest, name = line.split("  ", 1)
+        assert release.sha256_file(evidence / name) == digest
+
+
+def test_the_same_run_passes_when_nothing_skips(tmp_path):
+    passing = tmp_path / "test_passing.py"
+    passing.write_text("def test_x():\n    assert True\n")
+
+    report = release.run_gate(
+        _config(tmp_path),
+        output=tmp_path / "evidence",
+        stages=(("container", (str(passing),)),),
+        allowed_skips=release.CONTAINER_ALLOWED_SKIP_REASONS,
+    )
+
+    assert report["ok"] is True and report["unexpected_skips"] == []
+    assert report["revision"], "the evidence must say which commit it covers"
+
+
+# ── the command, and CI ──────────────────────────────────────────────────────
+
+
+def test_the_command_is_documented_and_wired():
+    from photo_pipeline.__main__ import main  # noqa: F401  (import proves it loads)
+
+    assert "container-gate" in (REPO / "photo_pipeline" / "__main__.py").read_text()
+    assert "container-gate" in (REPO / "README.md").read_text()
+
+
+def test_ci_runs_the_gate_on_main_and_keeps_its_evidence():
+    job = TEST_WORKFLOW["jobs"]["container"]
+    assert job["if"] == "gitea.event_name == 'push'", "pull requests have nothing to upgrade from"
+    script = "\n".join(step["run"] for step in job["steps"] if "run" in step)
+    assert "photo_pipeline container-gate" in script
+    evidence = next(step for step in job["steps"] if "upload-artifact" in str(step.get("uses")))
+    assert evidence["if"] == "always()", "a failed gate's logs are the ones worth keeping"
+    assert evidence["with"]["path"] == "gate-evidence"
+
+
+def test_the_upgrade_journey_can_still_reach_the_previous_version():
+    """Without depth, `git archive HEAD~1` has nothing to build."""
+    checkout = next(
+        step for step in TEST_WORKFLOW["jobs"]["container"]["steps"] if "checkout" in str(step.get("uses"))
+    )
+    assert checkout["with"]["fetch-depth"] >= 2

tests/integration/test_documentation.py

@@ -0,0 +1,179 @@
+"""US09-01: the documentation the application serves, checked without a browser.
+
+What is checkable offline is everything that makes the manuals *navigable* rather
+than merely present: every page reachable, every internal link and anchor landing
+somewhere real, the vendored renderer being the file that was pinned, and the
+application actually serving `docs/` in both the working copy and the image.
+
+The rendering itself — marked, mermaid, anchors, link rewriting — is proven in a
+real browser by ``tests/e2e/test_docs_ui.py``.
+"""
+
+from __future__ import annotations
+
+import hashlib
+import json
+import re
+from pathlib import Path
+
+from starlette.testclient import TestClient
+
+from photo_pipeline.api.app import create_app
+from photo_pipeline.api.security import DEFAULT_HEADERS
+from photo_pipeline.config import Config
+
+REPO = Path(__file__).resolve().parents[2]
+DOCS = REPO / "docs"
+VENDOR = REPO / "frontend" / "js" / "vendor"
+INDEX = DOCS / "index.md"
+
+MARKDOWN_LINK = re.compile(r"\[[^\]]*\]\(([^)\s]+)\)")
+HEADING = re.compile(r"^#{1,6}\s+(.+?)\s*$", re.MULTILINE)
+FENCE = re.compile(r"^```.*?^```", re.MULTILINE | re.DOTALL)
+
+
+def pages() -> list[Path]:
+    return sorted(DOCS.rglob("*.md"))
+
+
+def body(path: Path) -> str:
+    """A document without its fenced code, so an example link in a shell snippet is
+    not mistaken for a link the reader can follow."""
+    return FENCE.sub("", path.read_text())
+
+
+def slug(text: str) -> str:
+    """The heading-anchor rule, mirroring ``frontend/js/docs.js``.
+
+    Deliberately duplicated: this check has to run without a browser. The browser
+    test is the authority on the real behaviour, and it asserts the same anchors.
+    """
+    cleaned = re.sub(r"[^\w\s-]", "", text.lower(), flags=re.UNICODE).strip()
+    return re.sub(r"[\s-]+", "-", cleaned).strip("-") or "section"
+
+
+def anchors(path: Path) -> set[str]:
+    found: dict[str, int] = {}
+    for heading in HEADING.findall(body(path)):
+        # Markdown emphasis and inline code are not part of the rendered text.
+        plain = re.sub(r"[*`_]", "", heading)
+        base = slug(plain)
+        found[base] = found.get(base, 0) + 1
+    return {name if index == 1 else f"{name}-{index}" for name, count in found.items() for index in range(1, count + 1)}
+
+
+def internal_links(path: Path) -> list[str]:
+    return [
+        href
+        for href in MARKDOWN_LINK.findall(body(path))
+        if not re.match(r"^[a-z][a-z\d+.-]*:", href, re.IGNORECASE) and not href.startswith("//")
+    ]
+
+
+# ── the documentation tree ───────────────────────────────────────────────────
+
+
+def test_the_index_exists_and_names_every_page():
+    """A page nobody links to is a page nobody reads."""
+    assert INDEX.is_file()
+    listed = {
+        (INDEX.parent / href.split("#")[0]).resolve()
+        for href in internal_links(INDEX)
+        if href.split("#")[0].endswith(".md")
+    }
+    unreachable = [page.name for page in pages() if page != INDEX and page.resolve() not in listed]
+    assert unreachable == [], f"not linked from the index: {unreachable}"
+
+
+def test_every_internal_link_and_anchor_resolves():
+    broken: list[str] = []
+    for page in pages():
+        for href in internal_links(page):
+            target, _, anchor = href.partition("#")
+            destination = page if not target else (page.parent / target).resolve()
+            if not destination.is_file():
+                broken.append(f"{page.name} → {href} (no such file)")
+                continue
+            if anchor and anchor not in anchors(destination):
+                broken.append(f"{page.name} → {href} (no such heading)")
+    assert broken == [], broken
+
+
+def test_every_page_starts_with_one_title():
+    for page in pages():
+        titles = [line for line in page.read_text().splitlines() if line.startswith("# ")]
+        assert len(titles) == 1, f"{page.name} has {len(titles)} top-level titles"
+
+
+# ── the vendored renderer ────────────────────────────────────────────────────
+
+
+def test_the_vendored_libraries_are_the_files_that_were_pinned():
+    """A vendored dependency without a recorded checksum is a dependency nobody is
+    reviewing. Replacing one has to be a visible change to this manifest."""
+    manifest = json.loads((VENDOR / "VERSIONS.json").read_text())
+    recorded = {entry["file"]: entry for entry in manifest["libraries"]}
+
+    on_disk = {path.name for path in VENDOR.iterdir() if path.suffix in (".js", ".mjs")}
+    assert on_disk == set(recorded), f"unrecorded vendored files: {on_disk ^ set(recorded)}"
+
+    for name, entry in recorded.items():
+        digest = hashlib.sha256((VENDOR / name).read_bytes()).hexdigest()
+        assert digest == entry["sha256"], f"{name} does not match the pinned checksum"
+        assert entry["version"] in entry["url"], f"{name}: the pinned URL and version disagree"
+        assert entry["license"], f"{name}: no license recorded"
+
+
+def test_the_documentation_view_is_wired_into_the_shell():
+    shell = (REPO / "frontend" / "index.html").read_text()
+    assert 'data-nav="docs"' in shell, "the manuals are unreachable from the navigation"
+    assert '"/docs"' in (REPO / "frontend" / "js" / "app.js").read_text()
+
+
+# ── the policy the view runs under ───────────────────────────────────────────
+
+
+def test_the_script_boundary_is_unchanged_and_only_style_was_relaxed():
+    """The measured cost of rendering diagrams in the browser, held to that cost.
+
+    mermaid needs to style the SVG it builds, which `style-src 'self'` refuses. It
+    does not need to execute generated code, so `script-src` must never acquire the
+    escape hatch that would let injected markup run.
+    """
+    policy = DEFAULT_HEADERS["content-security-policy"]
+    directives = {
+        part.split(" ")[0]: part.split(" ")[1:]
+        for part in (piece.strip() for piece in policy.split(";"))
+        if part
+    }
+    assert directives["script-src"] == ["'self'"]
+    assert "'unsafe-inline'" in directives["style-src"]
+    for directive in ("default-src", "img-src", "connect-src", "font-src"):
+        assert "'self'" in directives[directive]
+        assert "'unsafe-inline'" not in directives[directive]
+    assert directives["frame-ancestors"] == ["'none'"]
+
+
+# ── serving it ───────────────────────────────────────────────────────────────
+
+
+def test_the_application_serves_the_markdown_without_a_session(tmp_path):
+    """Whoever cannot get past the access secret is exactly who needs these pages."""
+    config = Config.from_env({"PHOTO_PIPELINE_DATA_DIR": str(tmp_path / "data")})
+    (tmp_path / "data").mkdir()
+    with TestClient(create_app(config), raise_server_exceptions=False) as client:
+        response = client.get("/docs/index.md", headers={"cookie": ""})
+        assert response.status_code == 200
+        assert "Photo Pipeline documentation" in response.text
+        assert client.get("/docs/overview.md").status_code == 200
+        assert client.get("/docs/nothing-here.md").status_code == 404
+        # The mount is a directory, not a path parameter: nothing above it is reachable.
+        assert client.get("/docs/../pyproject.toml").status_code in (307, 404)
+        assert client.get("/docs/%2e%2e/pyproject.toml").status_code == 404
+
+
+def test_the_image_ships_the_documentation_it_serves():
+    """An image without `docs/` serves an empty manual — and the build context is
+    deny-by-default, so a new directory is excluded until it is named."""
+    assert "\n!docs\n" in (REPO / ".dockerignore").read_text()
+    assert "COPY docs ./docs" in (REPO / "Dockerfile").read_text()

tests/integration/test_installation_manual.py

@@ -0,0 +1,178 @@
+"""US09-02: the installation manual, checked against the thing it describes.
+
+Documentation rots quietly. A setting is renamed, a command grows a flag, an exit
+code changes meaning — and the manual keeps confidently saying the old thing until
+somebody follows it into a bad afternoon. So every fact in it that the code also
+knows is compared with the code, in both directions where both directions are
+meaningful: a setting the manual invents fails, and a setting the manual forgot
+fails too.
+"""
+
+from __future__ import annotations
+
+import re
+import subprocess
+import sys
+from functools import lru_cache
+from pathlib import Path
+
+from photo_pipeline.config import ENV_PREFIX, LEGACY_ALIASES, Config
+
+REPO = Path(__file__).resolve().parents[2]
+MANUAL = REPO / "docs" / "installation.md"
+MAIN = REPO / "photo_pipeline" / "__main__.py"
+
+TEXT = MANUAL.read_text()
+ENV_NAMES = re.compile(rf"{ENV_PREFIX}[A-Z0-9_]+")
+
+# Read by docker-compose.yml, not by Config. They are configuration an installer sets,
+# so the manual documents them; they are simply not application settings.
+COMPOSITION_ONLY = {
+    f"{ENV_PREFIX}IMAGE",
+    f"{ENV_PREFIX}LIBRARY_HOST_PATH",
+    f"{ENV_PREFIX}PUBLISH_ADDRESS",
+    f"{ENV_PREFIX}UID",
+    f"{ENV_PREFIX}GID",
+    f"{ENV_PREFIX}ENV_FILE",  # read before Config exists, so it is not a field
+}
+
+
+def documented_env_names() -> set[str]:
+    return set(ENV_NAMES.findall(TEXT))
+
+
+@lru_cache
+def cli_help(*args: str) -> str:
+    """What the CLI says about itself, asked the way an operator asks."""
+    result = subprocess.run(
+        [sys.executable, "-m", "photo_pipeline", *args, "--help"],
+        cwd=REPO,
+        capture_output=True,
+        text=True,
+        timeout=120,
+        check=False,
+    )
+    assert result.returncode == 0, result.stderr
+    return result.stdout
+
+
+@lru_cache
+def cli_commands() -> frozenset[str]:
+    listed = re.search(r"\{([a-z0-9,\-]+)\}", cli_help())
+    assert listed, f"the CLI listed no subcommands:\n{cli_help()}"
+    return frozenset(listed.group(1).split(","))
+
+
+def cli_flags(command: str) -> frozenset[str]:
+    return frozenset(re.findall(r"--[a-z][a-z-]+", cli_help(command)))
+
+
+# ── settings ─────────────────────────────────────────────────────────────────
+
+
+def test_every_setting_is_documented():
+    """A setting nobody documents is a setting nobody configures deliberately."""
+    expected = {f"{ENV_PREFIX}{field.upper()}" for field in Config.model_fields}
+    missing = sorted(expected - documented_env_names())
+    assert missing == [], f"undocumented settings: {missing}"
+
+
+def test_the_manual_invents_no_settings():
+    unknown = sorted(
+        name
+        for name in documented_env_names()
+        if name.removeprefix(ENV_PREFIX).lower() not in Config.model_fields
+        and name not in COMPOSITION_ONLY
+    )
+    assert unknown == [], f"documented but not a real setting: {unknown}"
+
+
+def test_the_documented_defaults_are_the_real_defaults():
+    """Spot-checked where a wrong default is actively dangerous: the bind address,
+    the trust boundary, and the gate that protects an irreplaceable library."""
+    assert Config.model_fields["host"].default == "127.0.0.1"
+    assert "`PHOTO_PIPELINE_HOST` | `127.0.0.1`" in TEXT
+    assert Config.model_fields["allowed_hosts"].default == ()
+    assert Config.model_fields["require_dry_run_approval"].default is False
+    assert "`PHOTO_PIPELINE_REQUIRE_DRY_RUN_APPROVAL` | `false`" in TEXT
+
+
+def test_every_secret_is_marked_as_one():
+    for field in ("access_secret", "vision_api_key", "immich_api_key"):
+        name = f"{ENV_PREFIX}{field.upper()}"
+        row = next(line for line in TEXT.splitlines() if line.startswith(f"| `{name}`"))
+        assert "secret" in row.lower(), f"{name} is not marked as a secret"
+
+
+def test_the_legacy_aliases_are_documented():
+    """An operator with an existing photo_analyzer.env needs to know it still works."""
+    for alias in LEGACY_ALIASES:
+        assert alias in TEXT, f"legacy alias {alias} is undocumented"
+
+
+# ── commands ─────────────────────────────────────────────────────────────────
+
+
+# The manual is an installation and operations manual, not a command reference: these
+# are the commands an installer actually runs, and every one of them must exist.
+OPERATIONAL = ("migrate", "serve", "worker", "backup", "verify-backup", "restore", "diagnostics")
+
+
+def test_every_command_the_manual_tells_you_to_run_exists():
+    missing = [name for name in OPERATIONAL if name not in cli_commands()]
+    assert missing == [], f"the manual names commands the CLI does not have: {missing}"
+    undocumented = [name for name in OPERATIONAL if not re.search(rf"\b{re.escape(name)}\b", TEXT)]
+    assert undocumented == [], f"operational commands the manual omits: {undocumented}"
+
+
+def test_every_documented_flag_exists_on_the_command_it_is_shown_with():
+    for command, flag in (
+        ("backup", "--reason"),
+        ("backup", "--keep"),
+        ("restore", "--into"),
+        ("worker", "--allow-legacy"),
+        ("serve", "--allow-legacy"),
+    ):
+        assert flag in cli_flags(command), f"{command} has no {flag}"
+        assert flag in TEXT, f"{flag} is undocumented"
+
+
+# ── exit codes ───────────────────────────────────────────────────────────────
+
+
+def test_every_documented_exit_code_is_one_the_cli_can_return():
+    documented = {int(code) for code in re.findall(r"^\| `(\d)` \|", TEXT, re.MULTILINE)}
+    returned = {int(code) for code in re.findall(r"^\s+return (\d)$", MAIN.read_text(), re.MULTILINE)}
+    assert documented, "no exit codes are documented"
+    assert documented <= returned | {0}, f"documented but unreachable: {sorted(documented - returned)}"
+
+
+def test_every_refusal_exit_code_is_documented():
+    """0 is success and 1 is 'it said why'; every other code is a specific refusal an
+    operator will meet at startup, and meeting an undocumented one is the worst case."""
+    returned = {int(code) for code in re.findall(r"^\s+return (\d)$", MAIN.read_text(), re.MULTILINE)}
+    documented = {int(code) for code in re.findall(r"^\| `(\d)` \|", TEXT, re.MULTILINE)}
+    missing = sorted(code for code in returned if code not in documented and code != 0)
+    assert missing == [], f"undocumented exit codes: {missing}"
+
+
+# ── the promises the manual makes ────────────────────────────────────────────
+
+
+def test_the_manual_carries_no_credential_shaped_example():
+    """The repository's own scanner refuses these in a commit; a manual is exactly
+    where a real-looking one gets copied from."""
+    suspicious = re.findall(
+        r"(?i)(api[_-]?key|secret|token|password)\s*[=:]\s*[\"']?([A-Za-z0-9_\-]{12,})",
+        TEXT,
+    )
+    assert suspicious == [], f"credential-shaped example: {suspicious}"
+
+
+def test_the_manual_states_the_invariants_an_installer_must_not_break():
+    for invariant in ("_IGNORE/", "One writer", "EXIF is verified before upload"):
+        assert invariant in TEXT, f"the manual does not state: {invariant}"
+
+
+def test_the_manual_is_reachable_from_the_index():
+    assert "installation.md" in (REPO / "docs" / "index.md").read_text()

tests/story_traceability.json

@@ -189,10 +189,26 @@
     ],
     "US08-04": [
       "tests/integration/test_deploy_workflows.py"
+    ],
+    "US08-05": [
+      "tests/integration/test_container_gate.py",
+      "tests/e2e/test_phase_h_container.py"
+    ],
+    "US09-01": [
+      "tests/integration/test_documentation.py",
+      "tests/e2e/test_docs_ui.py"
+    ],
+    "US09-02": [
+      "tests/integration/test_installation_manual.py"
+    ],
+    "US09-03": [
+      "tests/integration/test_architecture_overview.py",
+      "tests/e2e/test_docs_ui.py"
     ]
   },
   "planned": [
-    "US08-05"
+    "US09-04",
+    "US09-05"
   ],
   "_planned_comment": "Accepted backlog stories that are not implemented yet. The release gate (US07-07) requires every story file to be either mapped to tests or listed here, so an unimplemented story is a visible decision rather than a hole in the matrix."
 }