fix: close the findings the test suite surfaced

The suite that landed in #43/#44 was written against existing behaviour, so a
number of tests pinned bugs rather than asserting correct behaviour. This fixes
the production code and moves each of those tests onto the fixed behaviour in
the same change.

Security:

- project-download: derive the archive entry extension from the last path
  segment and restrict it to a short alphanumeric run, so an extensionless
  allowlisted url can no longer contribute a path separator; validate the r2
  branch against the strict proxy-path pattern instead of a `startsWith`, which
  let `/api/upload/video/clip.mp4/../../etc/passwd` through verbatim.
- rate-limit: hash a key or action wider than its column instead of skipping the
  query. Both the guard and the failing INSERT used to answer "allowed", so the
  limit stopped applying entirely. Warn at startup when TRUSTED_PROXY_MODE is
  unset in production.
- video uploads: the file name decides the content type; a client-declared video
  mime no longer makes `payload.exe` acceptable.
- email templates: escape in the helpers rather than relying on every caller,
  with an explicit `rawEmailHtml()` opt-out for the one call site that builds
  markup. `escapeHtml` now covers the single quote.
- CSP: allow loopback object storage outside production only.
- route-access: reach the billing redirect only for the workspace owner. Keying
  it off the owner's billing status alone made the redirect target an oracle for
  whose subscription had lapsed, and sent members to a page they cannot act on.
- search: carry the same billing condition every other read path carries.
- logger: check `err.name` as well as `err.constructor.name`, so a re-thrown,
  deserialised or minified Prisma error is still redacted.
- upload tokens: resolve the signing secret outside the try, so a server booted
  without one fails loudly instead of reporting every grant as a forgery.
- invitations: never downgrade an existing membership, and report a scoped
  invitation that points at nothing as not_found rather than accepted.
- auth: resolve the workspace role for every signed-in caller, so
  checkProjectAccess and computeProjectAccess stop disagreeing about the owner
  who also owns the workspace. The `intent` option is gone with it.
- r2-media-proxy: validate the object key inside the proxy so the guard travels
  with the function; delete the unused, unanchored `mediaUrlToR2Key`.
- r2: sign the content type into presigned PUT grants.

Correctness:

- frame rate snapping picks the nearest standard, not the first within
  tolerance, so 24, 30 and 60 fps are reachable at all.
- a version upload registers its Bunny cleanup as soon as bunny-init answers, so
  a failed tus upload no longer leaves a billed video behind.
- deleting videos clears storage before the rows, so a refused DELETE leaves a
  retryable row rather than an orphaned object.
- an expired upload session can be cancelled, which is what releases its quota.
- `voice/` joins the delete allowlist, so a voice note can be removed by the
  module that wrote it.
- a failed CORS write propagates instead of being mistaken for an empty config
  and replacing the bucket's rules.
- filtering projects by workspace no longer hides projects the unfiltered call
  returns.
- upload retries skip aborts and permanent 4xx; progress no longer divides by
  zero.
- reply edits no longer clear the comment's tag; optimistic resolve rolls back
  to the state it replaced; the delete snapshot is captured once.
- assorted UI fixes: duplicate React keys, double-click guards reading stale
  closures, the tag list fetched twice per load, a failed member list rendering
  as an empty one, a stale "Initializing upload..." beside a failure, and a
  registration banner pointing at an email that never arrives.

Consistency and access:

- the two download routes answer 404 for an id belonging to another tenant, as
  the comment export route already did. A caller who does belong still gets 403.
- accessible names for the share-link password field, the guest name gates, the
  version dialog inputs and the comment-tag controls.

Repository health:

- the runner image installs production dependencies only.
- a setup file for the unit project restores stubbed env centrally.
- native tsconfig path resolution replaces vite-tsconfig-paths.
- `uploadBytesWithProgress` exists once.
- admin stats bill Bunny storage to the workspace owner like every other
  quota, gate on the configured flag, wire up the single-flight guard and count
  the statuses that belonged to no bucket.
- `r2Client.destroy()` releases the presign client too.
- `prepare` tolerates a production install, where husky is absent.

tests/api/lib-project-access.test.ts

@@ -41,11 +41,6 @@ import {
   createWorkspace,
 } from '../factories';
 
-type Intent = 'view' | 'manage' | 'delete';
-
-/** `undefined` stands for a caller that passes no options at all. */
-const INTENTS: ReadonlyArray<Intent | undefined> = [undefined, 'view', 'manage', 'delete'];
-
 type Actor =
   | 'an anonymous caller'
   | 'an outsider'
@@ -252,42 +247,6 @@ const SCENARIOS: readonly Scenario[] = [
 ];
 
 // ---------------------------------------------------------------------------
-// The known divergence
-// ---------------------------------------------------------------------------
-
-/**
- * The one place the two functions disagree, spelled out rather than filtered
- * out.
- *
- * `shouldLoadWorkspaceRole` in lib/auth.ts skips the workspace queries for a
- * project owner on `intent: 'view'`, on the reasoning that an owner passes every
- * check on their own. That is true of the three permission booleans, and it is
- * why this is harmless today, but it is not true of the two identity flags:
- * checkProjectAccess() reports `isWorkspaceMember: false, isWorkspaceAdmin:
- * false` for an owner who is in fact the workspace owner, where
- * computeProjectAccess() on the same rows reports true and true.
- *
- * Harmless today rests on one fact and not on the design: the only consumer of
- * `isWorkspaceMember` from checkProjectAccess() is
- * app/api/versions/[versionId]/approvals/route.ts:37, and its
- * `isOwner || isProjectMember || isWorkspaceMember` is already satisfied by
- * `isOwner` for exactly the actor that diverges. Nothing consumes
- * `isWorkspaceAdmin` from checkProjectAccess() at all; `canEdit` folds it in,
- * and `isOwner` covers that too. The next reader of either flag inherits the
- * bug, which is why this is asserted instead of hidden behind a comparison of
- * `hasAccess` alone.
- *
- * If this stops matching, the divergence was closed: delete this function and
- * the test at the bottom of the file rather than widening either of them.
- */
-function knownDivergence(actor: Actor, intent: Intent | undefined): Partial<ExpectedAccess> {
-  const resolvesWorkspaceRole = intent === 'manage' || intent === 'delete';
-  if (actor === 'a project owner who also owns the workspace' && !resolvesWorkspaceRole) {
-    return { isWorkspaceMember: false, isWorkspaceAdmin: false };
-  }
-  return {};
-}
-
 // ---------------------------------------------------------------------------
 // Fixtures
 // ---------------------------------------------------------------------------
@@ -428,7 +387,7 @@ for (const scenario of SCENARIOS) {
     });
 
     for (const actor of ACTORS) {
-      it(`resolves ${actor} exactly as computeProjectAccess does, at every intent`, async () => {
+      it(`resolves ${actor} exactly as computeProjectAccess does`, async () => {
         const userId = seeded.userIdFor(actor);
         const projectId = projectIdFor(actor, seeded);
         const expected = scenario.expected[actor];
@@ -437,20 +396,10 @@ for (const scenario of SCENARIOS) {
         // The pure half first: given the rows, this is the answer.
         expect(computeProjectAccess(enriched, userId), 'computeProjectAccess').toEqual(expected);
 
-        // And the querying half, which has to arrive at the same place from the
-        // same rows, whatever the caller says it intends to do.
-        for (const intent of INTENTS) {
-          const checked = await checkProjectAccess(
-            enriched,
-            userId,
-            intent === undefined ? undefined : { intent }
-          );
-
-          expect(checked, `checkProjectAccess with intent ${intent ?? '(default)'}`).toEqual({
-            ...expected,
-            ...knownDivergence(actor, intent),
-          });
-        }
+        // And the querying half, which has to arrive at the same place from the same rows.
+        // It used to take an `intent` that skipped the workspace queries for an owner at
+        // `view`, which is what made these two disagree; there is one code path now.
+        expect(await checkProjectAccess(enriched, userId), 'checkProjectAccess').toEqual(expected);
       });
     }
   });
@@ -460,8 +409,13 @@ for (const scenario of SCENARIOS) {
 // The divergence, on its own
 // ---------------------------------------------------------------------------
 
-describe('checkProjectAccess and computeProjectAccess disagree in one place', () => {
-  it('hides the workspace role from a project owner who also owns the workspace, on intent view', async () => {
+describe('checkProjectAccess and computeProjectAccess agree about the workspace role', () => {
+  // This was the one cell of the matrix that diverged, and it is the shape every real
+  // signup produces: app/api/projects/route.ts gives a new project the workspace owner's
+  // id. checkProjectAccess() used to skip both workspace queries for an owner at `view`
+  // intent, so the two identity flags read as if this user were a stranger to the
+  // workspace they own.
+  it('reports a project owner who also owns the workspace as a workspace admin', async () => {
     const owner = await createUser();
     const workspace = await createWorkspace({ ownerId: owner.id });
     const project = await createProject({
@@ -471,32 +425,11 @@ describe('checkProjectAccess and computeProjectAccess disagree in one place', ()
     });
 
     const enriched = await fetchEnriched(project.id, owner.id);
-
     const computed = computeProjectAccess(enriched, owner.id);
-    const viewed = await checkProjectAccess(enriched, owner.id, { intent: 'view' });
-    const managed = await checkProjectAccess(enriched, owner.id, { intent: 'manage' });
 
-    // What the rows say: this user owns the workspace the project lives in.
     expect(computed.isWorkspaceMember).toBe(true);
     expect(computed.isWorkspaceAdmin).toBe(true);
 
-    // What checkProjectAccess() says on the intent that pages and GET routes
-    // use. `shouldLoadWorkspaceRole` skipped the workspace queries, so the role
-    // was never resolved and the two flags read as if this user were a stranger
-    // to the workspace.
-    expect(viewed.isWorkspaceMember).toBe(false);
-    expect(viewed.isWorkspaceAdmin).toBe(false);
-
-    // Ask the same question with a mutating intent and the same user, on the
-    // same rows, is a workspace owner again.
-    expect(managed.isWorkspaceMember).toBe(true);
-    expect(managed.isWorkspaceAdmin).toBe(true);
-
-    // Why nobody has noticed: every permission the flags feed is already
-    // granted by isOwner, so the divergence stops at the two identity flags.
-    expect(viewed.hasAccess).toBe(true);
-    expect(viewed.canEdit).toBe(true);
-    expect(viewed.canDelete).toBe(true);
-    expect({ ...viewed, isWorkspaceMember: true, isWorkspaceAdmin: true }).toEqual(computed);
+    expect(await checkProjectAccess(enriched, owner.id)).toEqual(computed);
   });
 });