An audit found tests that pass on broken code. The dominant pattern: fire a security
assertion at the all-zeros UUID and accept [403, 404] — the 404 comes from the resource
LOOKUP, not the guard, so the guard can be deleted and the test still passes. Repaired to
use real resources and demand exactly 403:
- banned-user cannot like / comment (the only coverage of those ban invariants)
- host cannot promote a real guest (or self) to admin — asserts nobody's role changed
- IDOR comment-delete already used a real resource; kept
Also inverted recovery.spec's "unknown name → nicht gefunden" test: that asserted the exact
account-enumeration oracle the F4 fix removed, so restoring the vuln would have made it
pass. Now: an unknown name must be byte-identical to a wrong PIN.
New coverage for paths that ran in production but in zero tests:
- quota.spec.ts: storage quota enforcement (413 over-limit, atomic increment under two
uploads held mid-body so both carry a stale total=0 — the real race; a naive Promise.all
version was itself vacuous and is documented as such). Proven to fail without the guard.
- authz-sweep.spec.ts: table-driven guest→403 / host→403 over ALL 19 privileged routes +
anonymous + __truncate. No live hole found; the whole surface is now locked.
- ban / unban / host-comment-delete AFTER release regenerate the keepsake (data-loss
paths that were dead under test); comment-delete mid-build doesn't strand the ZIP.
Lower-severity de-vacuuming: 10 MB comment test hit Caddy's 502 before the real 500-char
cap (now seeds a real upload, 501→400 / 500→201, mutation-verified); XSS name payloads
shortened under the 50-char cap so they actually store+render; ui-rendering XSS test now
proves the payload rendered before asserting no <b>; export page-object locators fixed to
the real "Download" label with a positive empty-state anchor; avatar palette spread test.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
lib/ conventions
Short rules. The patterns we already follow as of v0.16 — write new code that fits.
One store per cross-cutting concern. A single *-store.ts file owns each one:
auth.ts— JWT / PIN inlocalStorage,isAuthenticatedwritableui-store.ts— bottom-nav visibility, upload-sheet open state, FAB badge countdata-mode-store.ts— Saver vs Original media-loading preferenceprivacy-note-store.ts— admin-configured Datenschutzhinweis textquota-store.ts— live per-user storage snapshotupload-queue.ts— IndexedDB-persisted upload queue + processing state
Don't import these into other stores unless strictly necessary; let pages compose them.
DTOs mirror Rust types. All TS interfaces live in types.ts. Each one carries a
// mirrors backend/src/path::TypeName comment so the two stay searchable. If you add
a Rust DTO, add the TS twin in the same PR.
Gestures via Svelte actions in actions/. Long-press, double-tap, future swipe —
each is a use: action that fires a CustomEvent. Components stay free of gesture
plumbing.
Reusable bottom sheets via ContextSheet.svelte. Pass an actions: ContextAction[]
array. Any page that needs a long-press / kebab context menu uses the same primitive.
SSE relays are listed in sse.ts::KNOWN_EVENTS. New server event → add one entry
to that array, that's it.
Diashow transitions live in diashow/transitions/. Each is a Svelte component
plus one entry in transitions/index.ts. Adding a new animation is two-line work; no
diashow code needs to change.
No new global stores beyond the list above unless the new concept is genuinely app-wide. Page state belongs in the page.