- Keep the map footer links permanently visible; remove the hide toggle.
- Publish inline founder promote/demote from the member row.
- Add encrypted MLS reaction/delete/read control messages and UI.
- Add the minute Cordn join watcher: NIP-17 DMs + encrypted bell notices.
- Add the browser translation-matrix runner and a low-end performance finding.
- Add a browser-safe editorial generator with rotating satire and shipped highlights.
- Generate a fresh 1200x630 cover from the Drop #006 foundation and upload it to Blossom.
- Publish weekly drafts as NIP-78; add the admin Load draft/Preview composer flow and banner.
- Add the Monday user timer (draft by default) and a manual newsletter:weekly script.
- Remove static/live issue duplicates; translate the new admin strings to all 24 locales.
- @xenova/transformers 2.17.2 in a Web Worker; Opus-MT en↔es ONNX q8 packs
self-hosted on the relay VPS (/models + CORS); ONNX Runtime 1.14 WASM
self-hosted in public/ort/ (37MB) loaded from ../ort/.
- Hard network guard: only model GET/HEAD (no query) and /ort/ assets allowed.
Message text travels only via postMessage; 9 tests prove fetch is never called
with the text and every other request is blocked.
- TranslateMessage in DMs + crew chat; Settings auto-translate off by default;
memory-only translations; per-model download progress; 12 UI strings in all
24 locales.
- Both model directions verified from the hosted files; tsc, 138 tests, build
and smoke green.
- tinyld (MIT, lazy, local) + script rules + a weighted
Catalan/Galician/Basque/Spanish/Portuguese rule for the close/short cases
tinyld confuses. No network, no logs, no storage; conversation/contact prior
supported.
- Lingua with 24 languages produced a 99 MB WASM, so it was rejected as a
detector; tinyld + the gl rule covers all 24 at a fraction of the size.
- 28/28 language samples pass, including a test that fetch is never called.
- Phase 2: on-device Opus-MT (English pivot) in a worker, memory-only
translations, model packs on Blossom, sentinel/network-guard test.
- Dependency job is now 16 days (service env + script default); a separate
weekly scan-only timer reports npm/OSV findings without changing the lockfile.
- TimeoutStartSec=30min on the deps job and 20min on the deploy job (systemd's
90s default could kill a legitimate run); both scripts resolve the newest nvm
Node instead of hardcoding v22.23.1.
- npm override pins @tanstack/query-core to $@tanstack/react-query so a blanket
npm update cannot create a second copy (the Oct 26 tsc failure).
- Failed runs keep test/smoke logs + attempted package files in failures/
(rotated, 8 kept) and a last-failure marker; reports rotate (12 kept).
- scripts/release.sh: notes check -> npm test -> bump -> Android build ->
zsp publish -> relay verify -> commit/push + tag.
- Deploy/deps systemd units are tracked under scripts/systemd/.
- Full 8,703-key Hindi dictionary in src/i18n/hi (7 namespace batches),
placeholders verified, only 2% identical strings (brand/protocol names).
- Register hi in Language/LANGUAGE_LOADERS, detectLanguage (hi-IN),
LanguageToggle (🇮🇳, हिन्दी) and dateFnsI18n (hi-IN).
- Lazy-load Noto Sans Devanagari for hi via lib/localeFonts.ts (renamed from
cjkFonts.ts) and add it to the CSS font stack; other locales don't download it.
- check-i18n covers hi: 24 dictionaries / 25 codes at 100%; tsc/tests/build green.
- react-router-dom/react-router 7.18.4: the app only uses the v6-compatible API
(Link/NavLink/Routes/HashRouter + hooks), so zero code changes; tsc, tests and
the full-route boot smoke all pass. Clears the open-redirect advisory.
- vitest 5.0.5 (dev): clears vitest/@vitest/mocker/tinypool criticals. Setup now
imports @testing-library/jest-dom/vitest for the new expect types.
npm audit: 8 total (0 critical, 1 high = xlsx write-only, 7 moderate = build
tooling).
Bump 1.6.6 -> 1.6.7. Native Amber intents now send compressionType=none and the JS
signer decodes Signer1 gzip results, so long encrypted payloads no longer fail.
APK built + published to Zapstore (verified on relay.zapstore.dev).
Bump 1.6.5 -> 1.6.6, notes for the always-NIP-17 crew chat, instant sends +
persisted decrypt cache, the 20-locale i18n backfill, and the private-chat invite.
APK built + published to Zapstore (verified on relay.zapstore.dev).
Messages now appear on send; shared DM decrypt cache stops per-view re-decrypts;
double-send guarded. Bump 1.6.4 -> 1.6.5, APK built + published to Zapstore
(verified on relay.zapstore.dev).
Fix: sending to a crew sequential-ised its per-recipient signer intents, so Amber
approvals land instead of timing out. Bump 1.6.3 -> 1.6.4, APK built + published
to Zapstore (verified on relay.zapstore.dev).
Bump 1.6.2 -> 1.6.3, notes for the Amber navigation fix, the Messages Crews
section, and the crew encrypted-chat card rewrite. APK built + published to
Zapstore (verified on relay.zapstore.dev).
Bump 1.6.1 -> 1.6.2, release notes for the first-run checklist + backup-reminder
fixes. APK built and published to Zapstore (verified on relay.zapstore.dev).
- Crew page defaults to the Private-chat (MLS) tab when this device already has
the group (useCordnGroupExists reads the IndexedDB snapshot — local only, no
signer prompt); a manual tab click always wins.
- Founder/admin banner ('Set up') points at the tab when the crew has no group.
- Bump 1.6.0 -> 1.6.1 and add release notes for everything since 1.6.0 (private
crew chat/MLS, self-hosted relay + media, marketplace close-the-deal, trust
controls, map/weather, Clothing Designer, UI fixes).
Add src/lib/cordn/native/ — a browser-safe Cordn client doing real RFC 9420
MLS over a ContextVM coordinator, mirroring Cordn's wire contract (reimplemented,
not copied from @cordn/cli).
- mls.ts: identity, key packages, group create/add/join, application messages,
the ChaCha20-Poly1305 coordinator wrapper, causal prev tags.
- coordinator.ts: ContextVM/MCP-over-Nostr client on @contextvm/sdk, validated
with @cordn/core schemas.
- session.ts: CordnSession (key packages, createGroup, addMember->Welcome,
joinGroup, send, fetch) with persistence behind an injectable CordnStore
(MemoryCordnStore today; IndexedDB adapter TODO).
- Verified end-to-end against our own coordinator: a two-member round trip
(CORDN_LIVE=1 vitest run native.live.test.ts) — create, add, join, and
bidirectional encrypted messages with proven senders.
Deps: @cordn/core, @contextvm/mcp-sdk, @noble/ciphers pinned 2.2.0 (fixes the
previously invalid tree: ts-mls wants 2.2.0, a stale root copy was 0.5.3).
src/test/setup.ts is now guarded for node-environment test files. Not wired to
any UI yet; nothing imports the module, so the web bundle is unchanged.
Replace the license file with the unedited AGPL-3.0 text (copied in full
from the teambtcmap/btcmap-android LICENSE, byte-identical). No project
header and no 'or later' grant, so SPDX is AGPL-3.0-only. AGENTS.md records
the rationale and the section-13 source-offer implication.
Add the canonical AGPL-3.0 text at /LICENSE (with the project copyright
notice on top) and set package.json license to AGPL-3.0-or-later. Chosen
because the trust work reviews AGPL reference implementations and we
reimplement rather than copy; AGPL section 13 source-offer is covered by
the public repo + the License page's source link. Recorded in AGENTS.md.
Export ▾ now offers: PDF (full sheet), PDF (client quote — uses an optional
businessName), PNG summary card (canvas), XLSX multi-sheet workbook (lazy
SheetJS), CSV, Markdown, JSON, and Copy summary (clipboard).
Import auto-detects the CSV shape: a `name,order_cost,…` header is treated as a
supplier price list and replaces the Materials table; otherwise the CSV
round-trips the whole sheet. JSON still loads a backup.
The toolbar's primary Export now generates a real PDF (lazy-loaded jspdf) — a
formatted one-page summary: labour, equipment, operations, materials, processes,
packaging, margins and the selling prices, plus the date/currency. Print-to-PDF
was rejected because window.print() doesn't work in the Capacitor WebView.
A secondary JSON export stays (for Import/backup); Import still reads JSON.
jspdf is dynamically imported so it never enters the initial bundle.
Web NFC isn't exposed in Capacitor's WebView, so the APK needs a native writer.
Adds packages/skatespots-nfc — a local Capacitor Android plugin (Kotlin,
enableReaderMode + NDEF URL record), registered BY NAME (not bundled). lib/nfc.ts
now uses the plugin on native and Web NFC on the web. iOS can't write tags.
⚠️ UNVERIFIED (needs cap sync + a device). README documents it.
- First-visit language auto-detect (Spanish devices land in es; stored choice wins).
- Sasquatch AI replies in the app language (skate trick names kept English).
- i18n safety net: scripts/check-i18n.mjs wired into npm test — fails on any t()
key missing from en/es or any en/es key drift.
- ~1,680 new keys (2,297 total) across mapUi / spotUi / marketUi / labUi and
~65 files wired to t(): Map, spots (detail/create/edit/chat), marketplace,
trades, offers, escrow, shipping, The Lab, /become/* forms, coaches, skaters.
- Policy: only skate TRICK names stay English; user-generated content untouched.
packages/skatespots-background: a local Capacitor Android plugin that polls our
own push-service /pending and raises local notifications. Modes: off / standard
(WorkManager ~15 min) / fast (foreground service ~5 min). Kotlin +
AndroidManifest (POST_NOTIFICATIONS, FOREGROUND_SERVICE_DATA_SYNC, WAKE_LOCK,
battery-exemption) + Gradle module. Client: useBackgroundDelivery (registers a
poll token via NIP-98, seeds the plugin) + a Background delivery (Android)
control in NotificationSettingsCard. The app registers by NAME, not by importing
the package, so the web bundle is unaffected. UNVERIFIED - needs cap sync + a
device.
Server: /native/subscribe + /native/unsubscribe (NIP-98 proof), FCM device store,
and FCM HTTP v1 sending (service-account OAuth2, cached ~1h); /health reports
native + device count. 503 without PUSH_FIREBASE_SERVICE_ACCOUNT_JSON.
Client: @capacitor/push-notifications (lazily imported, native-only) in
usePushNotifications - request permission, register, store the FCM token, and
subscribe via /native/subscribe; tap deep-links into the app.
Needs a Firebase project: google-services.json + the Gradle plugin (android/),
and the service-account JSON in the host .env.
Our own keyless Bun service (not Linky's): watches relays for kind-1059 wraps
carrying the plaintext marker ["skatespots","push"] on the recipient copy,
extracts only {recipient, wrapId}, and sends a Web Push. NEVER decrypts - the
client fetches + decrypts the named wrap itself.
Endpoints: GET /health, GET /vapid-public-key, POST /auth/challenge ->
/subscribe|/unsubscribe with a one-time NIP-98 (kind 27235) challenge proof.
State is a JSON file; VAPID keys from env. Wired like the other standalone
servers (typecheck:push + check:push in npm test, deploy:push, Caddyfile +
systemd unit). Verified locally end-to-end: auth -> subscribe -> marked wrap ->
watcher -> push attempt, and proof replay rejected.
Client half (service worker, subscribe hook + Settings toggle, marker on DM
sends) is the next step.
Rebuilds the APK with everything shipped since 1.5.9 (all previously web-only):
destination search + bigger follow map, MintRadar ratings/reliability in a mint
detail sheet, counter-offers + barter with notifications and a Trades bell tab,
the crews type control, one-close spot sheet, and the floating map toolbar.
Rebuilt the APK so the phone gets:
- voice turn prompts (device TTS) + arrival, wake lock, bigger follow map
- floating map controls (map fills the screen)
Web was already deployed; this only bumps the Android version.
Rebuilt the release APK so the phone gets what shipped this session:
- Settings -> Coordinator (pick/add/health, run-your-own hand-off)
- crew form accepts a phone-run coordinator (nprofile/cordn link) + 'Use my coordinator'
- profile discovery: publish your coordinator, one-tap add on profiles
- Edit Profile preview + 'finish your profile' checklist
Web was already deployed; this only bumps the Android version.
An OS alarm an hour before each upcoming event the user RSVP'd going or
interested to. Because the OS holds the alarm it fires with the app fully
closed — the only closed-app notification achievable without a server.
- src/lib/reminders.ts: schedule/cancel via @capacitor/local-notifications,
tracking handed-out ids in localStorage so withdrawn RSVPs and passed events
are cancelled without disturbing one-off notifications. Web is a no-op.
- src/hooks/useLocalReminders.ts: reuses the existing upcoming-events and
my-RSVPs queries; no new relay destination, no signing.
- Bridge mounts the hook only on native and only when the new 'reminders'
pref is on; unmount cancels everything (opt-out/logout).
- Skipped the Pro-expiry reminder on purpose: it would drag the heavy
useSasquatchPro query into a nudge.
Bumps to v1.5.7 (versionCode 10507); APK rebuilt.
Rebuilt the release APK so the phone gets everything currently on main:
opt-in on-device notifications (incl. the @capacitor/local-notifications
native plugin), the batched/smoother map markers, profile badges, the
'closest spot to me' Sasquatch intent, and the 'Keep the map free' support
element. Bumps only package.json (version sync script derives versionCode).
Mirrors the activity the app already computes (bell, crews, DMs, newsletter)
to the OS notification shade, with no new relay traffic and no new network
destination.
- src/lib/osNotify.ts: one delivery API. Web uses the Notification API; native
uses @capacitor/local-notifications (lazily imported, so the web bundle never
pays for it). Tap-through navigates in-app.
- src/components/OsNotificationBridge.tsx: mounted once in Layout, inert until
opted in. Per-source silent priming + a seen-id set means history is never
replayed as a flood, and it stays quiet while the app is visible (the bell
already covers that).
- DMs notify from the gift-wrap ENVELOPE only (src/hooks/useDmEnvelopes.ts,
no decryption) — so Amber is never prompted just to be told you have a
message. Preview is honestly generic.
- src/lib/notificationPrefs.ts + useNotificationPrefs: device-local, opt-in,
default off, per-category toggles. Settings -> Notifications UI.
Deliberately local-only: a fully closed app needs a server push, which is not
built (Phase 2 = on-device background runner; Phase 3 = content-free FCM).
Bump 1.5.4 -> 1.5.5 and give today's ships their own release block: offline map
tiles, photo mirror fallback, the DM 'No messages yet' fix, and free
name@skate-spots.org handles. 1.5.4's notes are restored below.
APK rebuilt: versionName 1.5.5 / versionCode 10505, and its bundle carries the
ai.skate-spots.org base.
The package.json homepage pointed at skate-spots.com — a different domain on a
different host, not this app.
Also adds a deploy guard: auto-deploy.sh now hashes exactly what would be
uploaded (dist/ + netlify/ + netlify.toml) and skips the Netlify deploy when it
is byte-identical to what is already live, so metadata-only commits (version
bumps, homepage) stop costing 15 credits.
Same proxy as netlify/functions/ai.mjs, rebuilt as a standalone Bun service on
the CVM host so AI calls stop spending Netlify compute/bandwidth credits and
the sats-paid Routstr key lives on our box.
- ai-proxy/server.ts: Bun.serve on 127.0.0.1:8787 — NIP-98 + Pro gate, model
allowlist, per-IP throttle, /api/ai + /api/ai/status. Own tsconfig
(typecheck:ai in npm test); eslint-ignored like cvm-server.
- ai-proxy/{package.json,.env.example,skatespots-ai.service,Caddyfile,DEPLOY.md}.
- scripts/deploy-ai-proxy.sh (npm run deploy:ai): scp + bun install + unit +
restart, mirroring deploy-cvm.sh.
- Client: one place decides the AI base, src/lib/sasquatchAi.ts (also the NIP-98
u tag). Defaults unchanged (/api/ai on Netlify); set VITE_SASQUATCH_AI_BASE to
move it. Fully reversible.
Deployed and verified on the host: /api/ai/status returns the live Routstr
balance. Public TLS (DNS + Caddy) is the remaining step before switching the app
over, so no client behaviour changes yet.
SkateSpots now runs its own NIP-05 registry, so a new skater gets a human
handle from minute one instead of an npub.
- netlify/functions/nip05.mjs: /.well-known/nostr.json plus
/api/nip05/{check,claim,release}, backed by Netlify Blobs (first-party,
no new vendor). Claims are NIP-98 signed and written onlyIfNew, so only
the owner can claim a name for themselves. Platform names are seeded in
code and cannot be taken or released.
- src/lib/skateNip05.ts + src/hooks/useSkateNip05.ts +
src/components/SkateNip05Input.tsx: shared validation, debounced
availability, and the claim control.
- SignupDialog: a 'SkateSpots name' field, prefilled from the display
name. Claimed best-effort before the profile publish; a failure never
blocks signup, and the profile only advertises nip05 once confirmed.
- EditProfileForm: claim a free handle from the NIP-05 field.
Verified on a draft deploy: claim -> resolves -> a second key gets 409 ->
release -> gone. tsc/eslint/101 tests green.
1.5.3 shipped at 19:22, before the DM fix, the directions UI rework and the
newsletter reader fix. Bump to 1.5.4 so those can actually reach people.
- package.json 1.5.4 (-> versionName 1.5.4 / versionCode 10504).
- release-notes.txt: new v1.5.4 section (DM persistence + parallel send/read,
bigger-map/app-like directions, underscore italics in the issue reader);
trimmed v1.5.3 back to what was actually published.
1.5.2 is already published to Zapstore (verified: org.skatespots.app@1.5.2 on
relay.zapstore.dev), so the post-publish work needs a new version.
- package.json 1.5.3 (-> versionName 1.5.3 / versionCode 10503).
- release-notes.txt: new v1.5.3 section (real per-mode routing + steps + follow
mode + choose-your-start + offline saved routes + Amber scoping); trimmed the
v1.5.2 section back to what was actually published.
Bump package.json (single source of truth → versionName 1.5.2 / versionCode
10502 via scripts/sync-android-version.mjs) and add the v1.5.2 release notes:
marketplace deck ratings + seller dashboard, working in-app directions, nav
fixes (desktop More / mobile Marketplace), the native mints fix, and the
newsletter CPU + relay fixes.
Patch bump over the unreleased 1.5.0, folding in today's mobile fixes: Mints in
the phone nav, instant Directions, and no stray signer prompt for the AI balance.
Sasquatch Pro (AI), in-app directions (transit/bike/moto/car), Barbers &
Hairstylists in The Lab, and a smoother app shell (no remounts, no stray
signer pop-ups).
- scripts/deploy-cvm.sh: backup remote server.ts -> upload -> restart -> confirm
the heartbeat, with --rollback. Host/SSH values come from the never-committed
~/.config/skatespots/cvm.env (or CVM_HOST), so no host details ever enter this
public repo.
- package.json: add `npm run deploy:cvm`.
- AGENTS.md: record that the CVM server runs on ONE external host (deploy via
deploy:cvm) and that the discovery relay, not the server, is the SPOF.
Load protection (public read-only endpoint — protects relays + this box):
- Per-pubkey sliding-window throttle and in-flight cap via inbound middleware.
The transport can't return a JSON-RPC error, so over-budget callers are dropped
and time out (logged).
- Global semaphore caps concurrent upstream relay queries.
- Short-TTL cache keyed by filter set, so repeated identical queries (e.g. a
burst of get_community_stats) don't re-hit the relays.
- All tunable via CVM_RATE_* / CVM_MAX_* / CVM_QUERY_CACHE_MS (documented in
.env.example); maps pruned by a janitor.
CI:
- cvm-server/tsconfig.json type-checks the standalone server; new
typecheck:cvm / check:cvm scripts, with typecheck:cvm wired into npm test so
the server can't rot unnoticed (it was excluded from tsc + eslint).
Data-driven pass (adds rollup-plugin-visualizer behind ANALYZE=1):
- MapPage was statically imported by the router, dragging the map route AND the
IndexedDB offline-cache layer into the entry -> now lazy()
- NewsletterOverlay imported NEWSLETTER_ISSUES from the whole NewsletterPage
module -> types + data moved to lib/newsletterIssues.ts
- SasquatchGuide (assistant overlay) lazy-loaded in Layout
- Cashu stack is now a separate chunk (useCashu), off the critical path
Net: entry down ~61%; first paint no longer parses the map, the newsletter
page, the assistant or the wallet stack.
- Regenerate lockfile so @tanstack/react-query (and the related
query-core / persist-client packages) resolve to 5.102.8, matching
package.json. The committed lockfile was stale at 5.75.1, which made
`npm ci` abort with "lock file's @tanstack/react-query@5.75.1 does not
satisfy 5.102.8" and left vitest running against a half-broken
node_modules.
- Use path.resolve(__dirname, './src/test/setup.ts') for vitest
setupFiles so the setup module is located relative to the config file
rather than the invocation cwd, fixing the "Cannot find module
.../src/test/setup.ts" resolution failure.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Two type errors (surfaced by `npm run test` / tsc --noEmit on a full
toolchain; the in-browser bundler does not run tsc, so they were latent):
- useCVMClient.ts: remove a malformed cast
`spots as typeof spots & { tags; created_at }[]` feeding
deduplicateByIdentity(). The intersection dropped pubkey/kind and broke
the reassignment to `spots`. deduplicateAddressable() already returns
the correct shape, so the cast was both wrong and unnecessary.
- queryPersister.ts: import PersistedClient / Persister from
@tanstack/query-persist-client-core (their canonical source) instead of
@tanstack/react-query-persist-client, whose `export *` re-export tsc
won't resolve for named type imports. Added the core package (5.102.8,
matching the pinned wrapper) as an explicit dependency.
Both are pre-existing issues on the branch, unrelated to the audit
remediation commits.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Three UX wins so the app feels "already loaded" the moment it opens:
1. IndexedDB React Query persistence (lib/queryPersister.ts)
- PersistQueryClientProvider rehydrates the last-seen spots, shops,
events, tours, crews, coaches, listings and author profiles from
IndexedDB on cold start, so returning users land on a populated map
instead of a blank one; fresh data refetches quietly in the
background.
- Strict allow-list / deny-list: only public, JSON-serializable,
cold-start-relevant queries are persisted. Private/encrypted data
(stash, DMs, contacts, wallet/nwc/cashu/bitcoin, signer) is NEVER
written to disk. 24h maxAge + version buster keep it fresh & safe.
- Pinned @tanstack/react-query + query-core + persist-client to a
single 5.102.8 to avoid a query-core mismatch (partialMatchKey).
2. Optimistic geolocation (lib/lastLocation.ts)
- getCurrentPosition now caches every successful fix.
- The map opens ALREADY centered on the user's last-known area
(city zoom) instead of the world view; "you are here" + proximity
sort seed from the cache instantly.
- "Near Me" flies to the cached location immediately on tap, then
refines to the fresh GPS fix. Silent background warm-up on web map
mount so a later tap resolves instantly.
3. Cold-start skeleton (components/MapSkeletonPins.tsx)
- When the map is genuinely empty + loading, show elegant pulsing
ghost pins + a "Finding spots" pill instead of a blank map; they
vanish the instant real (or cached) spots render. Splash now
signals content-ready when spots actually resolve.
Respects prefers-reduced-motion throughout. Build + runtime verified
clean (no console errors).
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Prevents the android:sync script from regressing versionCode/versionName
back to 1.3.7. 1.4.4 → versionCode 10404 (above the previous 10403).
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
package.json pinned nostr-tools to an exact 2.13.0 while the committed
package-lock.json resolved 2.15.0 (root spec ^2.13.0), so `npm ci` on a
local machine refused to run ("lock file's nostr-tools@2.15.0 does not
satisfy nostr-tools@2.13.0"). Relaxed the pin to ^2.13.0 to match the
lockfile and every other dependency's caret range. Cloud build verified.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Two greenlit onboarding improvements, keeping the existing layout intact.
#5 Save-your-login banner (the safety one):
- Right under the green "You're in!" bar, logged-in users who haven't backed up
see a prominent amber banner — "Save your login — do this now" — that taps
straight to Security settings. It reads the shared `backup-confirmed` flag
(same source of truth as BackupReminder), so it PERSISTS until they actually
back up; unlike the dismissible reminder sheet, it can't silently turn into
"never". Once confirmed it flips to a calm green "Login saved" note. Plain
language ("save your login", not "back up your nsec"). Pulse dot respects
reduced-motion.
#1 Lighten first-glance load:
- The 9-item FAQ ("Common questions") is now a collapse-by-default section — one
tidy row until tapped — via a small reusable CollapsibleSection (aria-expanded,
reduced-motion aware). Nothing removed; the page just feels lighter on arrival.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Full self-audit of the Citrine feature. Fixes found:
Robustness
- Probe now tries BOTH ws://127.0.0.1 and ws://localhost and routes to whichever
actually answers (probeCitrineEndpoints + CitrineProbeResult.url). We no longer
bet the whole feature on one loopback address form; useCitrine.url reflects the
winning endpoint so reads/writes go to the right place.
Correctness
- Preview toggle was mutating the REAL setting while in dev-preview; handleToggle
is now a no-op under preview (forced state owns the visuals).
Accessibility / "flawless" bar
- Respect prefers-reduced-motion: status pips, chip pop-in and the expand slide
now motion-reduce:animate-none (no looping glow/pulse for reduced-motion users).
- Recheck button gets an aria-label; status text is role="status" aria-live so
screen readers announce searching→detected transitions; decorative pips are
aria-hidden.
UX copy
- "Waiting" state no longer presumes Citrine isn't installed. Clearer two-step
guidance (Open Citrine → Start; come back → Recheck) covering the common
"installed but not started/detected yet" case.
Honest note: MIXED_CONTENT_ALWAYS_ALLOW (1.3.5) is the high-probability fix for
the 0-connections block but is not device-verified from here; the dual-endpoint
probe hedges the second most likely cause (address form).
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
The real fix for "Citrine running but 0 connections": the app is served over
https in the WebView and Citrine listens on plaintext ws://127.0.0.1, which
modern Android WebViews block as mixed content BEFORE it reaches the network
layer — allowMixedContent + the loopback network-security-config don't fully
cover WebSocket mixed content.
MainActivity now sets the bridge WebView's mixed-content mode to
MIXED_CONTENT_ALWAYS_ALLOW after super.onCreate(), which lets the localhost
ws:// connection through for every user. All real traffic stays wss:///https://;
this only unblocks the loopback relay. The WebViewClient is left untouched
(overriding it previously caused a black screen). Wrapped in try/catch so it can
never crash the launcher.
Also removed the temporary on-screen probe debug readout (kept the diagnostic
fields in useCitrine, unused by UI, for future use).
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Citrine is running (listening on ws://127.0.0.1:4869) but the app can't connect
and shows 0 connections — a WebView-layer block, not our logic. To diagnose
without adb/USB, the probe now returns { ok, reason } capturing which stage
failed (blocked / connect-error / timeout-no-open / opened-but-no-nostr-reply /
closed-before-open with code), and the Settings card shows a small debug readout
(probe url + reason) whenever not connected. Purely additive; no behaviour
change to routing.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Self-audit of the 1.3.0–1.3.2 Citrine feature found two real issues:
1. False-positive detection (the important one). probeCitrine resolved true on
bare socket OPEN, so ANY process squatting on 127.0.0.1:4869 would be treated
as Citrine — and we'd then route reads AND published events at a non-Nostr
peer. The probe now completes a minimal Nostr handshake: open → send a bounded
REQ with a random sub id → require a protocol-shaped reply (EVENT/EOSE/CLOSED
echoing the sub id, or a NOTICE) before declaring the relay live. Non-JSON /
non-Nostr responses fall through to a clean timeout=false. A real relay
answers with EOSE in ~ms on localhost, so the happy path is unchanged.
2. Probe churn. The reachability query re-opened a fresh ws:// probe every 5 min
and on every window focus for the whole session — needless battery/wakelock,
and a localhost poke on every Amber round-trip. Now it only polls (30s) and
refetches-on-focus WHILE still waiting for Citrine to appear; once connected
it stops the background churn (a per-session local relay won't vanish
silently, and a failed read re-probes naturally).
Reviewed and left as-is (deliberate, safe): writing every kind to Citrine incl.
gift-wrapped DMs / Stash is intended on-device encrypted backup — all sensitive
kinds are NIP-44 encrypted before reaching the publish layer, so nothing
plaintext-sensitive is mirrored. Preview flag remains render-only.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
The CitrineCard is Android-only (renders null off-native), so its four visual
states couldn't be reviewed without an APK build. Adds a strictly cosmetic
preview switch:
localStorage.setItem('citrine:preview', '1') // show card, real behaviour
localStorage.setItem('citrine:preview', 'live') // force connected state
localStorage.setItem('citrine:preview', 'waiting') // force not-detected
localStorage.setItem('citrine:preview', 'searching') // force searching
localStorage.removeItem('citrine:preview') // back to normal
Safety: the flag unlocks RENDERING ONLY via isCitrinePossible(). It never sets
citrine.active and never yields a relay url, so no query or publish is ever
routed anywhere different — production relay behaviour is identical with or
without the flag. A forced state is purely visual (enabled/probing are derived
locally in the card for preview; the real probe still governs active/url).
lib/citrine.ts: citrinePreview()/isCitrinePreview() + CitrinePreviewState.
useCitrine: exposes `preview`. CitrineCard: forced state overrides the computed
state for review only.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Completes the local-relay integration so it actually works end-to-end and feels
first-class.
Native (Android):
- android-templates/network_security_config.xml — permits cleartext for
LOOPBACK ONLY (127.0.0.1 / localhost / ::1); all real traffic stays TLS. This
is what lets ws://127.0.0.1 (Citrine) connect on API 28+, which otherwise
silently blocks cleartext.
- scripts/inject-network-security.mjs — idempotently copies the XML into
res/xml/ and adds android:networkSecurityConfig to <application>, mirroring
the existing inject-amber-queries pattern so it survives `cap sync`.
- Wired into android:init / android:sync / android:build.
Reads AND writes now reach Citrine app-wide:
- NostrProvider's NPool routers (reqRouter/eventRouter) fold in the active
Citrine URL via a ref, so every default-pool query hits it first (instant
local reads) and every useNostrPublish write also lands on-device (backup +
offline capture). Previously only explicit useSkateRelays paths benefited.
- A dedicated effect refreshes ['nostr'] queries when the local relay comes/goes
so reads switch over immediately. Fully no-op off-Android (url stays null).
UI/UX (Relays settings CitrineCard):
- Four crisp states — off / searching / live / waiting — with a breathing then
glowing status pip, benefit chips (Instant reads · Works offline · On-device
backup) that pop in on connect, a spinning Recheck, an actionable install CTA,
and a persistent encryption-reassurance note.
- Graceful first-run: shows "searching" until the first probe settles (via new
useCitrine.settled), never flashing the orange "not detected" state.
Tailwind: added status-breathe / status-live / pop-in keyframes+animations.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Integrates Citrine (github.com/greenart7c3/Citrine), an Android app that runs a
full Nostr relay locally on the phone. Because it's "just a relay URL", it folds
into the existing useSkateRelays / nostr.relay() machinery with no new protocol.
What it buys SkateSpots:
- Instant reads: a localhost hit (~1ms) is queried first, remote relays run in
parallel and reconcile — every skate-content surface feels faster.
- Offline-in-the-field: reads work with no signal; writes land on-device and
publish to remote when connectivity returns.
- Resilience/backup: a durable local copy of the user's own events survives
public-relay outages and addressable-event eviction.
- Privacy: local-first data can lean on the device instead of public relays.
Design / safety:
- Android-only (Capacitor native + platform === 'android'); renders nothing and
never probes localhost on iOS/desktop web. Mixed-content ws:// is already
permitted via capacitor allowMixedContent.
- Opt-in only via config.useCitrineRelay (default off). No data is ever routed
through Citrine unless the user enables it AND a relay is actually reachable.
- Reachability probe (short-lived WebSocket) cached through react-query, shared
app-wide, re-checked periodically + on focus, with a manual Recheck button.
- When active, the local URL is prepended to reads and included in writes; the
read-relay set feeds getQueryKeySuffix so caches refetch automatically when
Citrine comes/goes.
- Encryption unchanged: Stash/DMs stay NIP-44 encrypted regardless of
destination — the Settings card states this explicitly. Local ≠ plaintext.
New: src/lib/citrine.ts, src/hooks/useCitrine.ts, CitrineCard in the Relays
settings screen. AppConfig gains useCitrineRelay (zod-validated, optional).
On-device verification still needed: confirm Android permits ws://127.0.0.1
cleartext to localhost (network security config) once Citrine is running.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
NIP-17 group messages (rumor with >1 recipient p-tag) were bleeding into the
1:1 surfaces:
- useConversation (thread view): a group message from `otherPubkey` where I was
also a recipient satisfied isFromOther && isToMe, so group chatter appeared
inside the 1:1 thread.
- useConversations (list): group messages were collapsed to conversationWith =
sender (or an arbitrary recipient for my own sends), spawning phantom 1:1
entries and double-listing every group (once here, once in
useGroupConversations).
Both 1:1 hooks now skip any rumor with more than one p-tag. Groups remain the
sole responsibility of the useGroupConversation(s) hooks, which already require
pTags.length > 1. Genuine 1:1 rumors carry exactly one recipient p-tag
(createRumor doesn't tag the sender), so no legitimate message is dropped.
Verified solid (no changes): the seal(13)/rumor(14) kind checks, the
rumor.pubkey === seal.pubkey impersonation guard, and the amberSafeQueryOptions
decrypt-storm protection for interactive signers are all correct.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
The composer derived the kind 30023 addressable `d` identifier straight from
slugify(title). Because 30023 is addressable, two issues sharing a slug —
a recurring "Weekly Roundup", or "Update!" vs "Update?" — would share a
coordinate and the newer publish would silently overwrite the older issue.
Blank slug now auto-generates a unique coordinate (title-slug + UTC date +
short base36 time suffix). An explicit manual slug is still honored verbatim,
so intentionally editing/replacing a specific issue keeps working.
Verified clean (no changes): backfill panel is idempotent (skips slugs already
on-relay, deterministic issue-{n} identifiers) and the subscriber-count hook
uses the standard newest-marker-per-author tally with no double-counting.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Audited the spot detail surface — status reports, ownership, and
community contributions. It's well-built:
STATUS (useSpotStatus) — a proper trust×recency consensus: each report
weighted by recencyWeight (30-day half-life) × trustWeight (0.2 floor so
everyone counts, trusted users dominate), only the newest report per
reporter counts (no ballot stuffing), reports past 180 days are dropped,
and the parser rejects any unrecognised status value so garbage can't
poison the tally. Single + batch paths both invalidate on publish.
CONTRIBUTIONS — photos (kind 1063) and tag suggestions (kind 1111) are
SEPARATE events referencing the spot via #a, so contributors never touch
the owner's addressable spot event (no clobber). Tag votes dedupe per
contributor. The detail gallery spreads owner images first so the hero
is always the owner's photo when present; community shots are badged.
OWNERSHIP — transfer/claim model is intentional (claiming republishes
under the claimer's own key = their own coordinate; the original
survives, as the UI states). One real rough edge: handleClaim copied ALL
original tags then appended claim-from/claim-reason, so claiming a spot
that was itself a prior claim would STACK multiple provenance pairs on
the event. Fixed by filtering out any existing claim-from/claim-reason
before appending a single fresh pair pointing at the immediate prior
owner. No functional/authorization change.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Audited the wallet surface (real money: Cashu proofs, Lightning, NWC).
Most of it is solid — the Cashu singleton init with browser-CORS mint
probing/auto-heal, the send/receive token flows, the Lightning receive
path (double-tap guard + redeemed-quote listener cleanup + actionable
CORS errors), and the melt-based send all check out and match the zap
path.
One real money-loss gap: removing a mint calls
mintRepository.deleteMint(), which drops that mint's proofs from the
device — so removing a mint you still hold sats in strands those funds.
There are two remove-mint dialogs and they disagreed:
• MintPrivacyManager already warned with the concrete at-risk balance.
• CashuWallet's dialog showed only a generic static "make sure you've
withdrawn funds" line — no balance, no escalation — so a user could
delete a mint holding sats with effectively no warning.
Brought CashuWallet up to (slightly past) parity: when the mint being
removed has a non-zero balance it now shows a loud red warning with the
exact sats at risk and relabels the button "Remove anyway"; empty mints
keep the mild yellow note. Deletion is still allowed (a permanently
offline mint must be removable even with a stuck balance) — it's now an
informed choice, not a silent one.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Audited all three remaining surfaces. Two of them needed no changes:
MARKETPLACE / GEARSWAP — sound. useMarkListingSold preserves every tag
and flips only `status` (author-gated, right comment explaining why it
doesn't rebuild). Queries validate + dedupe, exclude showcase-only Stash
items from the sellable market, and GearSwap filters out expired/sold
listings and separates WTB from WTS. No bugs.
STASH — sound, and notably careful. The private list is one NIP-44
encrypted-to-self blob; all three mutations (add/update/remove) use
fetchLatestStash() (fetch-before-write) which falls back to non-empty
cache and THROWS rather than risk clobbering the user's gear, and is
Amber-gated (staleTime: Infinity). The public/private showcase
reconciliation in StashPage is complete: going public publishes the
kind-30402 showcase event, and going private / deleting / converting to
a wishlist item all call unpublishItem — so a public event can't linger
after an item is made private (no privacy leak). No bugs.
THE LAB — the collective join-request clobber was already fixed in
1.2.37. One robustness gap remained: useLabProfile fetched the profile
with limit:1 and trusted whatever came back. Relays can hold multiple
revisions of an addressable event, so a stale revision could win —
showing an old bio or, worse, a roster missing recently-approved
members. Now fetches a few (limit:10) and reduces to the newest
created_at, matching the pickMostRecent discipline already in useCrew.
On detail pages an author filter is always set, so newest-of-one-author
is exactly right.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Audited the whole crew surface. The trust model is genuinely solid:
membership is double-sided and unforgeable (founder p-tags + the
member's own NIP-32 kind-1985 attestation, cross-referenced in
useCollectiveMembers), roles are founder-controlled via the crew's
addressable event (useCrewRole), moderation only honours labels authored
by listed moderators, and the invite/join/endorse flow is symmetric with
proper author validation and batched, deduped queries throughout.
One real bug found — a stale-snapshot member clobber. The crew roster is
a single addressable event rewritten wholesale on every member add, but
three write sites rebuilt it from the CACHED crew/profile snapshot
without re-fetching first:
• useApproveJoinRequest (founder "Add to crew")
• useSendCrewInvite (founder pre-list on invite)
• useApproveCollectiveJoinRequest (Lab collective owner add)
So approving two join-requests back to back rebuilt the second write from
the pre-first-write roster and silently DROPPED the member added by the
first (addressable events replace by coordinate). This is exactly the
fetch-before-write discipline already applied to spots/events/shops/
contacts/stash — the crew/collective roster writes were the ones missed.
Fix: added fetchLatestCrewEvent() (and an inline equivalent for Lab
profiles) that re-queries the freshest roster event by coordinate and
merges the new member into THAT, never regressing to an older revision,
falling back to the passed-in snapshot only on a relay hiccup so an add
still goes through. Founder-only / owner-only authorization and the
membership-attestation trust model are unchanged.
Member removal / admin promotion go through the founder's own profile
editor (a deliberate full-roster overwrite by the sole owner), so they
aren't concurrent-merge scenarios and are left as-is.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Two real bugs surfaced auditing the Map page.
1. The "trusted-first" soft sort was silently dead for the default
config. filteredSpots ran a standalone .sort() to float trusted
content up, then IMMEDIATELY ran a second unconditional .sort() by
sortBy — which re-ordered everything and wiped the trust ordering.
For a logged-in user with ≥5 follows the smart default is
trustFilter='trusted' + sortBy='recent', so the documented
"sort trusted first, badge unverified" behaviour never actually
happened. Fixed by folding the trust tier into the MAIN comparator as
its primary key, with the chosen sortBy as the tiebreaker within each
tier. Strict modes (verified/following) still hard-filter as before.
2. MiniStashCard (gear pins in the map drawer) built its border colour
with an interpolated Tailwind class (`border-${colour}-500/40`) that
the JIT compiler can't see, so it was never generated — dead class
emitting an invalid string into the DOM. The inline
style={{ borderColor }} that actually worked was also applied
unconditionally, overriding the intended cyan border on WISHLIST
items with the rarity colour. Now: wishlist keeps its static cyan
border; gear uses the rarity colour via inline style only.
Geolocation (native vs web, denied/timeout/no-provider handling),
batched trust/status/comment/weather queries, and the layer plumbing
into SpotMap were all audited and are sound.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Audited the RSVP / event-update notification flow end to end. The data
layer was already solid and secure — RSVPs p-tag the host, updates
p-tag every RSVPer, and useActivityNotifications guards event-update
spoofing by requiring the author to be the event's host (from the `a`
coordinate) or an admin. The gap was on the rendering side.
Fixes:
- ActivityNotificationCard hardwired every notification to link to the
ACTOR's profile. For an RSVP ("Alice is going to your event") the host
wants the event + attendee list, and for an event update ("moved to
Sunday") the attendee wants to READ the update — not visit a profile.
Both events carry the event's `a` coordinate (31923:pubkey:identifier),
so these two types now route to /event/:pubkey/:identifier. Everything
else still lands on the actor's profile. d-tags containing ':' are
reconstructed correctly.
- useRsvp only invalidated the single-event RSVP query, so the grid /
calendar "going" counts lagged behind by up to the 30s staleTime after
RSVPing. Now also invalidates the batch-event-rsvps query so counts
update promptly.
- Social-tab empty-state copy omitted RSVPs and host updates even though
they land there; made it honest.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Events and tours could only be browsed as a card list — there was no
way to see what's happening on a given date, which is exactly the
clarity the calendar ask called for.
New EventsCalendar component: a purpose-built month grid (not the
date-picker) with:
- Prev/next/today month navigation
- Per-day item dots (max 3 + overflow count) with correct multi-day
overlap handling for ranged events and competition tours
- A selected-day agenda panel showing title, time (or date range /
"All day" for tours), location, and RSVP going counts, each linking
straight to the detail page
- Month total footer, empty-day state, full responsiveness, and
aria-labels / aria-pressed for accessibility
Wired a List⇄Calendar layout toggle into EventsPage for both the
events and tours tabs. Calendar mode reuses the already-filtered
event/tour sets (search, trust, type filters all still apply), hides
the list-only date-window chips, and shows a matching skeleton while
loading. Tours without a concrete date range are omitted from the
calendar (they have no fixed date) and remain browsable in list view.
Going counts are threaded through from the existing batch RSVP query,
so the calendar adds no extra relay load.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
Tours (kind 30823) were the only major addressable entity with no edit path: an
organizer had to delete and recreate (losing the URL/identifier) just to fix a
typo or price. Rather than duplicate the enormous multi-category tour form (and
risk schema drift / the exact tag-clobber bugs fixed earlier this session), I
made the existing CreateTourPage do double duty:
- CreateTourPage now accepts an optional `existing?: SkateTour`. When present it
prefills every field (all four categories: local/national/international/
destruction — including challenges, prize distribution, film-release, partner
shop, cities/countries, coords, dates), reuses the same `d` identifier, and
adapts copy (title, submit button, toast).
- On save it PRESERVES unmanaged tags (sponsors, tags from other clients) and
keeps any existing paid `featured` window — same "preserve unmanaged tags"
discipline as the event/shop edit fixes, since 30823 is addressable and
re-publishing replaces the whole tag set.
- Guarded the category→duration reset effect so it no longer clobbers the
prefilled duration on the first render in edit mode.
- The paid FeaturedUpgrade upsell is hidden when editing (existing status is
carried through automatically).
New EditTourPage wrapper loads the tour, gates to the owner (login / access-
denied states), and renders <CreateTourPage existing={tour} />. Added the
/edit-tour/:pubkey/:identifier route and an owner-only Edit button on
TourDetailPage beside the existing Delete control.
This closes the one concrete gap the addressable-events audit surfaced: every
entity now has both a working edit path (preserving unmanaged tags) and owner
management.
Co-authored-by: shakespeare.diy <assistant@shakespeare.diy>
