The send_webhook automation step fetched an account-controlled URL with
attacker-controlled headers and no SSRF protection, following redirects
by default — letting an agent+ reach cloud metadata (169.254.169.254),
loopback, and RFC1918 hosts from the server's network (status code
surfaces in automation logs). The separate webhook_endpoints delivery
path already guards this; the automation path did not.
Route send_webhook through the existing isDeliverableUrl guard before the
fetch, and mirror the deliver.ts hardening: redirect: 'manual' (a public
URL can't 3xx-bounce to an internal one) and a 10s timeout. Add an engine
regression test asserting a link-local target never reaches fetch.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Migration 029 created claim_ai_reply_slot(uuid, integer) as SECURITY
DEFINER but never GRANTed EXECUTE — the only function in the schema
missing its grant (007/012/017/018/019/025/030 all grant theirs).
SECURITY DEFINER sets the privileges a function runs *with*, not who may
*call* it: the caller still needs EXECUTE. On instances where the default
PUBLIC execute privilege on public-schema functions has been revoked
(hardened / self-hosted Supabase), the service_role that PostgREST uses
cannot invoke it. The AI auto-reply path runs entirely under the
service-role client (the inbound webhook has no auth.uid()), so
db.rpc('claim_ai_reply_slot') fails permission-denied, the dispatch bails
before sending, and the bot silently never answers ANY inbound message.
The Playground works because it never claims a slot — matching the report.
- 031_ai_reply_slot_grant.sql: new migration granting EXECUTE to
service_role (029 is already applied on existing installs, so it needs
its own migration to reach them). service_role only — the sole caller —
matching the 007/012 increment-counter precedent.
- 029_ai_reply.sql: add the same grant next to the definition so fresh
installs are correct.
- auto-reply.ts: log a real claim RPC error instead of returning
silently, so this class of deploy failure is diagnosable.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Promotes the AI assistant out of Settings into a first-class "AI Agents"
sidebar area.
- Sidebar: new "AI Agents" item → /agents.
- /agents page with two tabs:
- Playground — a test chat (POST /api/ai/playground) that runs the
real auto-reply path (KB retrieval + provider, auto_reply prompt,
multi-turn, shows handoff). loadAiConfig gains requireActive so you
can test before enabling.
- Setup — the existing provider/key + knowledge-base config, relocated.
- Removed the AI section from Settings (settings-sections + page); same
account config, new home. No migration.
Tests +6 (query, loadAiConfig requireActive). 593 pass; typecheck, lint,
build green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Security: REVOKE ALL FROM PUBLIC + GRANT EXECUTE (authenticated,
service_role) on the match RPCs so anon can't read a KB by supplying
an account_id (mirrors migrations 018/025).
- Ingest: store chunks even when embedding fails (insert first, rethrow
after) so a doc stays lexically searchable; the "lexical still works"
warning is now truthful.
- Retrieval: short-circuit on an empty KB (cheap COUNT) to avoid a paid
query embedding + 2 RPCs on every draft/auto-reply.
- Migration: HNSW instead of IVFFlat (robust on small/growing tables);
language-neutral 'simple' FTS config (was 'english'); semantic RPC
takes text and casts ::vector(1536) to remove PostgREST bind ambiguity.
- Embeddings key: distinguish corrupt-from-absent (loadEmbeddingsKey
returns {key,corrupt}); log on decrypt failure; reindex now reports
failure instead of false success; POST/PATCH warn on keyword-only save.
- embedTexts: fail loud when the provider omits result indices (was
silently misaligning chunks with vectors).
Tests: +3 (empty-KB guard, embed-failure-still-indexes, missing-index).
587 pass; typecheck/lint/build green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Grounds the AI assistant (0.5.0) in the account's own content so it can
answer instead of handing off.
- Hybrid retrieval: Postgres full-text search for every account; optional
semantic search (pgvector + OpenAI text-embedding-3-small) when an
embeddings key is set. Semantic-primary, topped up with FTS.
- src/lib/ai: chunk, embeddings, knowledge (ingest + retrieve), query;
buildSystemPrompt gains a knowledge block; config carries an
(optional) embeddings key.
- Wired into /api/ai/draft and the auto-reply bot (best-effort — never
blocks a reply).
- KB API (/api/ai/knowledge CRUD + /reindex) and a Settings → AI
Assistant knowledge-base manager + embeddings-key field.
- Migration 030: pgvector, ai_knowledge_documents, ai_knowledge_chunks
(fts + embedding), match RPCs, ai_configs.embeddings_api_key.
Migration required: supabase/migrations/030_ai_knowledge.sql.
Stacked on feat/ai-reply-assistant (#329).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds an AI reply assistant powered by the account's own OpenAI or
Anthropic key (stored AES-256-GCM-encrypted; no per-seat AI fee).
- AI-drafted replies in the inbox composer (✨ button, agent+), via
read-only POST /api/ai/draft.
- Optional auto-reply bot on inbound messages that no Flow consumed and
no agent owns, with a per-conversation cap and clean [[HANDOFF]] to a
human. Stands down when a message-level automation would also reply.
- Settings → AI Assistant (admin+): provider/model, key, business
context, toggles, and a Test-key button that verifies before saving.
- Provider layer (src/lib/ai) with OpenAI + Anthropic adapters behind
one interface; migration 029 (ai_configs + per-conversation columns +
atomic claim_ai_reply_slot RPC).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Completes the public API roadmap (#245): register HTTPS endpoints to be
POSTed to when things happen in an account, so automations can react to
inbound activity instead of polling.
- Migration 028: account-scoped webhook_endpoints (secret stored
AES-GCM-encrypted) + an atomic record_webhook_failure() function.
- Scope webhooks:manage; management API GET/POST /api/v1/webhooks and
GET/PATCH/DELETE /api/v1/webhooks/{id} (secret returned once).
- Events message.received / conversation.created / message.status_updated
dispatched from the inbound webhook handler's after() block.
- Signed delivery (X-Wacrm-Signature = HMAC-SHA256 over t.body), single
best-effort attempt, auto-disable after consecutive failures.
- Versioned 0.4.0; docs/public-api.md Webhooks section; CHANGELOG.
Fixes from the adversarial code review of this diff:
- SSRF: block private/loopback/link-local (incl. cloud metadata) targets
at delivery time + redirect:'manual' so a public URL can't bounce
internal (src/lib/webhooks/ssrf.ts).
- Status mirror UPDATE no longer uses .select().maybeSingle() (message_id
is non-unique); account resolved via a bounded limit(1) read.
- Atomic failure counter via SQL RPC (was a racy read-modify-write).
- conversation.created emitted before the reaction short-circuit.
- Per-delivery id added for receiver dedupe; docs state at-least-once /
unordered delivery.
- Signature verify tolerates hex case / header whitespace.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Complete the public API roadmap on top of the existing key-auth layer
(#290). Re-lands POST /api/v1/messages (lost when the stacked #291 never
reached main) by extracting the dashboard send route's core into
`sendMessageToConversation`, then adds:
- Contacts: GET/POST /api/v1/contacts, GET/PATCH /api/v1/contacts/{id}
- Conversations: GET list + /{id} + /{id}/messages
- Broadcasts: POST /api/v1/broadcasts + GET /{id} (persist rows, fan out
in after(), poll for status)
All list endpoints share one keyset-pagination contract
({ data, meta: { next_cursor } }). No migration required — scopes and
tables already exist. Versioned as 0.3.0 (folded into [Unreleased]);
docs/public-api.md updated; outbound webhooks left as the sole roadmap
item. API stays v1 (additive).
Includes fixes from an adversarial code review: cursor validation
against filter injection, broadcast counts deferred to the DB aggregate
trigger, terminal broadcast status reflects failures, duplicate-phone
de-dup, up-front send-param validation (no orphan conversations),
PATCH null-clear semantics, error-checked tag sync, bounded tag filter,
unified audit-user attribution, and a maxDuration on the broadcast
fan-out.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds contact-based filters to the Inbox conversation list (issue #272):
- Multi-select tag filter with OR logic (a conversation matches if its
contact has any selected tag), consistent with Broadcast audiences.
- Company filter (exact match) over companies present in the inbox.
- Filters stack with the existing status filter and search; active
selections show as removable chips with a clear-all control.
Tags are hydrated via a shared CONVERSATION_SELECT embed and normalized
onto contact.tags, so the realtime hydrate path keeps new conversations
matching an active filter. Filtering logic lives in a pure, unit-tested
matchesContactFilters helper.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
getCurrentAccount() and the client AuthProvider resolved the caller's
account through a PostgREST embedded join (`account:accounts!inner(...)`)
on profiles. That embed requires PostgREST to have the
profiles.account_id -> accounts.id relationship in its schema cache.
When the cache is stale -- a common Supabase state right after a
migration adds the FK, or when migrations are applied out of band -- the
embed fails hard with PGRST200 and the whole account context collapses.
It is the only query shape in the app that depends on relationship
inference (everything else reads accounts with plain `.eq('id', ...)`
filters), which is why the dashboard kept working while Settings ->
API keys / Team members returned 403 "Could not load account context"
and the Overview's cards hung loading (issue #294).
Replace the embed with a plain account-by-id lookup in both the server
helper and the client AuthProvider. Same accounts RLS, no relationship
inference, graceful degradation. The client now also keeps account_id /
account_role even if the account-name lookup itself fails. Adds a
regression test for getCurrentAccount.
Fixes#294
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds the machine-to-machine auth layer the public REST API needs so
external scripts and automations can drive wacrm (issue #245). This is
the groundwork release; data endpoints land one at a time on top of it.
- Migration 026: account-scoped `api_keys` table. SHA-256 hash at rest
(plaintext shown once), display `key_prefix`, `scopes[]`, RLS via
`is_account_member` (members read, admin+ manage).
- `src/lib/api-keys/`: key gen/hash (wacrm_live_ prefix, timing-safe),
scope vocabulary, service-role auth-path store.
- `requireApiKey(request, scope?)` (src/lib/auth/api-context.ts):
bearer parse -> hash lookup -> liveness -> per-key rate limit ->
scopes-only authz. Returns a service-role-scoped account context.
- `/api/v1` envelope ({data} / {error:{code,message}}) + GET /api/v1/me
reference endpoint that exercises the whole stack.
- Dashboard: admin+ key management routes + Settings -> API keys tab.
- Docs (docs/public-api.md), CHANGELOG, README, .env note.
Checks: typecheck clean, lint 0 errors, full suite green (+27 tests).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds a live presence layer so the Settings → Team members roster and the
inbox Assign dropdown show who is actively using the dashboard, idle, or
gone. Implements #269.
How it works
- New `member_presence` table + `touch_presence` SECURITY DEFINER RPC
(migration 024). Account-scoped SELECT RLS mirrors the read-only
roster; the RPC derives account_id server-side, never client-supplied.
- The active client heartbeats 'online'/'away' every ~30s; 'offline' is
derived by viewers from staleness, so a closed tab resolves to offline
without an unreliable unload write.
- Pure, unit-tested derivation/formatting in src/lib/presence.ts.
- `usePresence` reads the table, subscribes to Realtime (subscribe-then-
merge so events landing mid-fetch aren't clobbered), and re-derives
offline on a 15s tick.
- Roster: avatar badge + tooltip + "N online · N away · N offline"
summary, with an aria-label so AT/keyboard users get presence too.
Inbox Assign dropdown: matching dot before each teammate.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Creating a keyword_match automation with Match type left on the default
"Contains" failed activation with `match type must be "exact" or
"contains"` (path trigger.match_type). The dropdown showed "Contains"
via a display-only fallback, but the value was never written to the
config, so the saved payload omitted match_type entirely.
Both engines already treat a missing match_type as "contains"
(automations/engine.ts, flows/engine.ts), so the activation validator
was stricter than runtime. Two coordinated fixes:
- validate.ts: accept a missing match_type (defaults to "contains");
only an explicit, unrecognised value is rejected. Aligns the
validator with the engines and any non-UI client.
- automation-builder.tsx: seed match_type: "contains" on mount of
KeywordMatchConfig so the displayed default is actually persisted,
keeping the saved record explicit and self-describing.
Adds a regression test that a keyword_match config with no match_type
passes activation validation.
Fixes#271
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Creating a template with an image header failed because the code sent
`example.header_url` — Meta only accepts a Resumable-Upload media handle
(`example.header_handle`) at creation time, never a plain URL.
- Resumable Upload: new `uploadResumableMedia` (meta-api) implements the
two-step `/{app_id}/uploads` flow → media handle. Needs META_APP_ID.
- Handle derivation: `ensureImageHeaderHandle` fetches the header image
(uploaded file OR pasted URL) and turns it into a handle before submit.
Wired into both the submit and edit (PATCH) routes with actionable 400s
(missing META_APP_ID, unreachable URL, wrong type, >5 MB). This fixes
BOTH the new upload path and the previously-broken manual-URL path.
- Template builder: add an image file upload (account-scoped chat-media
bucket via uploadAccountMedia) with preview + JPEG/PNG + 5 MB validation,
alongside the URL field.
- Send-time fix: an approved image template now sends the public
header_media_url as `link`; the resumable header_handle is NO LONGER
used as a send media id (it's invalid for sends and would fail).
- META_APP_ID documented as required for image-header templates.
Scope: image headers (the issue). Helper is generic for video/document.
typecheck/lint/build green; 421 tests (10 new).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Photo/video/document/voice send has been validated in production, so it's
now available to every account. Removes the 'chat_media' beta gate:
- Composer always shows the attach menu + voice recorder (no
useBetaFeature check).
- Send route no longer 403s media kinds for non-opted-in accounts.
- Deletes the now-unused beta-features helper, the useBetaFeature hook,
and their test (added solely for this rollout; no other consumer).
The profiles.beta_features column (migration 011) is left in place for
future betas. typecheck/lint/build green; 411 tests pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Follow-up to #259. Voice notes are encoded to Ogg/Opus in the browser via
opus-recorder (small WASM encoder, lazy-loaded on record), so there's no
server-side transcode and nothing to provision on the host. The server
stores the .ogg like any other attachment; WhatsApp renders it as a voice
note.
- Composer: MediaRecorder + /api/whatsapp/transcode-audio replaced with
opus-recorder. Encoder worker vendored at public/opus/encoderWorker.min.js
(no install-time download to fail), ignored by eslint + prettier.
- Removed the transcode route, the audio-transcode lib + tests, the
ffmpeg-static dep + ambient types, the FFMPEG_PATH env doc, and the
unused transcodeAudio rate limit.
- Added opus-recorder ambient types.
Still behind the 'chat_media' beta flag with per-kind size caps.
typecheck/lint/build green; 414 tests pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Media sends (photo/video/document/voice) are now opt-in per account via
the existing profiles.beta_features mechanism (migration 011 — same
pattern used to dogfood Flows before GA). No migration needed.
- New shared `hasBetaFeature` + `CHAT_MEDIA_BETA` ('chat_media') key, and
a `useBetaFeature` client hook mirroring useCan (false while loading).
- Composer: the attach menu, hidden file inputs, and voice recorder only
render when the flag is on; the hint indent reverts so non-beta users
see the unchanged text + template composer.
- Server: the send route rejects media kinds (403) and the transcode
route rejects (403) unless the caller's profile is opted in — so the
endpoints can't be driven ahead of the rollout. Text + templates stay
available to everyone.
Enable for one account via Supabase Studio:
UPDATE profiles SET beta_features = array_append(beta_features, 'chat_media')
WHERE email = 'someone@example.com';
Tests: hasBetaFeature null-safety + membership (419 total). typecheck/lint/build green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Addresses senior-review findings on the media-send feature:
- Storage lifecycle: uploads are GC'd on the abandonment paths
(discarded draft, failed/errored send, unmount with a staged draft,
and replacing one draft with another) via a new deleteAccountMedia
helper. Previously every cancel/failure orphaned a public object.
- Per-kind size caps (image 5MB, video/audio/document 16MB) mirror
Meta's limits and are enforced BEFORE upload, so a too-large image no
longer uploads then fails at the Meta call.
- Voice notes auto-stop at a 5-min cap and always render as a voice note
(transcode anything that isn't already OGG/Opus — Safari MP4 included),
not just Chromium WebM.
- Send route rejects unknown message_type and captions over Meta's 1024
cap up front; transcode route guards its output size too.
- DraftPreview hoisted to a stable module-scope component so the caption
input no longer remounts (and drops focus) on each keystroke.
Tests: per-kind cap assertions added (418 total). typecheck/lint/build green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Agents can now attach a photo/video/document or record a voice note in
the inbox composer and send it live in a 1:1 thread — previously media
could only be received from customers or sent via the Flows builder.
- 023_chat_media.sql: new account-scoped `chat-media` Storage bucket
(mirrors flow-media RLS from 020) plus Meta-accepted audio MIME types.
- meta-api: sendMediaMessage gains an `audio` kind (link-only — Meta
rejects caption/filename on audio; WhatsApp renders OGG/Opus as a
voice note).
- /api/whatsapp/send: media branch reusing the phone-variant retry loop.
- /api/whatsapp/transcode-audio: Node-runtime route remuxing the
WebM/Opus that Chromium's MediaRecorder produces to OGG/Opus via
ffmpeg-static (Firefox/Safari record accepted formats and skip it).
- composer: attach menu, file pickers, voice recorder + auto-transcode,
preview with caption; thread: optimistic handleSendMedia.
- shared uploadAccountMedia helper (flows form refactored onto it).
- next.config: allow microphone=(self) (was fully blocked) + CSP
media-src; document optional FFMPEG_PATH override.
Tests: meta-api audio payload shape, buildMediaPath, audio-transcode
helpers (14 new). typecheck/lint/build green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The sweep over-tokenized the neutral status badges in the three status
data modules to bg-muted/border-border, which broke
broadcast-status.test.ts ("each variant has the dark-theme class
triple") — it asserts every badge is bg-*/10 + text-* + border-*/20.
The badge system intentionally uses fixed palette colors at low opacity
(translucent fills are mode-tolerant). Restore the original slate fills
(so dark is identical and the triple contract holds) but keep
text-muted-foreground for the label, so neutral statuses stay legible in
light mode where a solid slate-400 would be too faint on white.
- broadcast-status: draft / recipient pending
- template-status: DRAFT / PENDING_DELETION
- trigger-meta: time_based / fallback
All 400 tests pass.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Introduce a light/dark MODE axis orthogonal to the existing accent
themes. Neutral surface tokens now live in html[data-mode="light|dark"]
blocks; accent tokens (--primary*, --ring, --chart-1, --sidebar-primary*)
stay in html[data-theme] blocks, so any accent works in either mode.
- globals.css: split neutrals (mode) from accent (theme); add light
palette; dark values unchanged (dark stays the default).
- lib/themes.ts: Mode type, DEFAULT_MODE=dark, MODE_STORAGE_KEY, isMode.
- use-theme.tsx: provider owns mode/setMode/toggleMode + cross-tab sync.
- layout.tsx: boot script applies persisted mode before first paint
(no flash); colorScheme "dark light".
- mode-toggle.tsx (header sun/moon) + themed-toaster.tsx (mode-aware).
- appearance-panel.tsx: light/dark control + updated copy.
- Tokenize hardcoded neutral colors (slate/white/gray) -> semantic
tokens across the app so every surface flips with mode. Flow canvas
and message-bubble timestamps made mode-aware too.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Reformatted code in `import-modal.tsx` for better readability, including consistent indentation and line breaks.
- Enhanced `parse-contact-csv.ts` and `parse-contact-csv.test.ts` with consistent string delimiters and improved whitespace handling.
- Updated comments for clarity and consistency across the import functionality.
- Ensured consistent use of single quotes for strings throughout the files.
- Updated `parseContactCsv` to include a flag for the presence of a company column in the CSV.
- Modified the import modal state management to handle the new company column, improving user experience during contact import.
- Enhanced the `ImportPreviewTags` component to conditionally render based on the presence of company data.
- Adjusted the UI to better display tag and company information in the import preview.
- Added `ImportPreviewTags` component to display tags with color coding in the import modal.
- Updated `parseContactCsv` to return an object containing both parsed rows and a flag indicating the presence of a tags column.
- Modified state management in the import modal to handle tags and their colors, improving user experience during contact import.
- Adjusted the import modal to conditionally render the tags column based on the presence of tags in the preview.
- Introduced `parseContactCsv` for improved CSV parsing, replacing the previous inline parsing logic.
- Added support for importing tags alongside contacts, allowing users to assign tags during the import process.
- Implemented `assignImportedContactTags` to handle tag assignments after contact creation.
- Updated the import result state to include the number of tags assigned and added relevant success and info notifications for users.
- Enhanced the import modal description to clarify the format for tag input in CSV files.
The "Update Contact Field" automation action could only write the built-in
name/email/company columns. This wires up the existing custom_fields /
contact_custom_values tables so automations can read and update arbitrary
custom fields, and adds the missing UI to define them.
- Add a "Custom fields" manager dialog on the Contacts page (admin+ gated)
to create / rename / delete field definitions. Inserts carry account_id.
- Load account custom fields into the automation builder's ResourcesContext
and offer them in the Field dropdown, encoded as `custom:<id>`. A saved
field that's since been deleted is preserved as an "(unknown field)"
option so editing doesn't silently drop it.
- Engine: interpolate the value ({{ vars.* }} / {{ message.text }}) for all
fields, and for `custom:<id>` verify the field belongs to the automation's
account (defense in depth) before upserting into contact_custom_values.
- Tests: cover the custom-value upsert, variable interpolation, and
cross-account rejection.
Closes#235
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The same phone could become multiple contacts: only the WhatsApp
webhook de-duped, while manual create and CSV import inserted raw
un-normalized phones, and the DB had no uniqueness. That fragmented
conversations, deals, and tags across duplicates.
- Migration 022: generated `contacts.phone_normalized` (digits-only),
a one-shot `merge_duplicate_contacts()` that collapses existing
dupes into the oldest row (re-pointing conversations, deals, notes,
tags, custom values, broadcast recipients — no data loss), then a
UNIQUE (account_id, phone_normalized) index as the authoritative
guarantee for every write path.
- Shared `src/lib/contacts/dedupe.ts`: findExistingContact (suffix
prefilter + phonesMatch), isExactMatch, isUniqueViolation,
dedupeByPhone. The webhook now reuses findExistingContact (and
recovers from a unique-violation race instead of dropping the
message).
- Contact form: on-blur duplicate check — hard-block + "view existing"
link on an exact match, soft warning on a trunk-variant near-match;
unique-violation backstop on submit.
- CSV import: de-dupes within the file and against existing contacts,
reports "imported / skipped / failed", treats 23505 as skipped.
- Contact type gains account_id (was missing) + phone_normalized.
- Unit tests for the dedupe helpers.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
deals.currency has no DB CHECK (only accounts.default_currency does),
so legacy rows, imports, or hand-edited data can hold a structurally
invalid code (e.g. "United States"). Intl.NumberFormat throws a
RangeError on such codes, which would crash any deal-card / contact
render that passes deal.currency straight through.
Guard formatCurrency with a try/catch that falls back to "CODE 1,234"
so a bad value degrades gracefully instead of taking down the view.
Add currency.test.ts locking in the defaults, the no-throw contract,
and formatCurrencyShort's symbol/fallback behavior.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
wacrm hardcoded USD across every deal-value formatter, the new-deal
form, and automation-created deals, making deal tracking unhelpful for
non-US self-hosted users. Add a per-account default currency.
- Migration 021: accounts.default_currency TEXT NOT NULL DEFAULT 'USD'
with a 3-letter-code CHECK. The existing admin-only accounts_update
RLS policy already gates writes.
- New src/lib/currency.ts: single source of truth (CURRENCIES list,
formatCurrency, formatCurrencyShort), replacing ~6 duplicated
USD-hardcoded helpers.
- useAuth exposes defaultCurrency (joined from accounts), falling back
to USD while loading / on pre-021 schemas.
- Settings → Deals tab to pick the currency (admin-only, read-only for
others to match RLS).
- Deal form defaults new deals to the account currency and sources its
dropdown from CURRENCIES; existing deals keep their saved currency.
- Pipeline board/analytics, dashboard "Open Deals Value", and the
pipeline donut format totals in the account currency. Individual
deal cards still show each deal's own currency.
- Automation create_deal sets the new deal's currency from the
account default instead of the static DB default.
One currency per account (no FX conversion) per the issue's
recommended first pass.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The automation engine executes steps via the service-role client, which
bypasses RLS. POST /api/automations/engine reads contact_id straight from
the request body, so an authenticated attacker could supply another
tenant's contact UUID and have the engine modify it (update_contact_field),
tag it, or read it back through a condition — bypassing the contacts RLS
policy. (GHSA-63cv-2c49-m5v3)
- Primary: verify the supplied contactId belongs to the dispatching
account before any step runs; refuse silently otherwise.
- Defense in depth: scope the service-role contacts writes/reads
(update_contact_field, contact_field condition) by account_id. Tag
steps rely on the entry-point guard since contact_tags has no
account_id column — documented inline.
- Add engine.test.ts covering the guard (PoC inverted), the happy path,
and the account-scoped write.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Addresses the L1 finding from the final security review. The six
admin-class write endpoints under /api/account had no rate
limits, so a compromised admin session (or a script run in a
loop) could spam invite creation / role flips / member removes
fast enough to flood the DB or audit logs.
Adds a new `RATE_LIMITS.adminAction` bucket — 30/min/user — and
wires it into:
- POST /api/account/invitations → admin:inviteCreate
- DELETE /api/account/invitations/[id] → admin:inviteRevoke
- PATCH /api/account/members/[userId] → admin:memberRole
- DELETE /api/account/members/[userId] → admin:memberRemove
- POST /api/account/transfer-ownership → admin:transferOwnership
- PATCH /api/account → admin:rename
Each endpoint keys its own bucket so spamming one doesn't starve
another. 30/min is comfortably above any realistic human pace
(the Members tab is a clicks-only UI; a legit admin would
struggle to hit ten of these per minute by hand) while still
bounding accidental abuse.
The bucket joins the existing send / broadcast / react /
invitationPeek / invitationRedeem buckets in `RATE_LIMITS` —
same in-memory fixed-window limiter, same `rateLimitResponse`
helper for the 429 shape.
Severity in the final review: LOW. The admin-trust model means
this isn't blocking — but the cost of adding it is six
near-identical 4-line additions and zero net risk, so worth
landing for hardening.
Verification
- `npm run lint` → 0 errors (same 19 pre-existing warnings)
- `npm run typecheck` → clean
- `npm test` → 366/366 pass
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PR 8 of the multi-user accounts series. Flips the WhatsApp webhook,
automations engine, and flows engine from per-user tenancy to
per-account so a shared inbox actually behaves like one.
Why this matters
- Pre-this-PR: an inbound message resolved the recipient by
whatsapp_config.user_id and dispatched only that user's
automations / flows. Teammates joining via an invite would see
the message in the shared inbox but their automations / flows
never fired.
- Also fixes incipient NOT NULL violations from migration 017.
That migration added `account_id NOT NULL` to automation_logs,
automation_pending_executions, flow_runs, and deals. Inserts in
the engines hadn't been updated yet — any production run would
500 with a column violation. This PR adds account_id to every
such insert.
Surface
- `src/app/api/whatsapp/webhook/route.ts`
- Resolves `config.account_id` AND `config.user_id` from the
matched whatsapp_config row. Passes accountId through every
downstream call. user_id is kept as `configOwnerUserId` and
used only for NOT NULL FK audit columns on contact /
conversation inserts.
- `findOrCreateContact` / `findOrCreateConversation` look up by
account_id and insert with both columns.
- `flagBroadcastReplyIfAny` matches broadcasts by account_id so
any teammate's broadcast is updated when a customer replies.
- `dispatchInboundToFlows` and `runAutomationsForTrigger` now
take `accountId` instead of `userId`.
- `src/lib/automations/engine.ts`
- `DispatchInput.userId` → `accountId`.
- `runAutomationsForTrigger` queries automations by account_id.
- `executeAutomation` inserts into automation_logs with both
account_id (tenancy) and user_id (audit, from automation row).
- `wait` step inserts into automation_pending_executions with
account_id.
- `create_deal` step inserts into deals with account_id.
- `assign_conversation` / `close_conversation` /
`resolveConversationId` filter by account_id.
- `resumePendingExecution` row shape now includes account_id;
cron passes it through.
- engineSendText / engineSendTemplate calls pass
`accountId: args.automation.account_id`.
- `src/lib/flows/engine.ts`
- `DispatchInboundInput.userId` is kept (audit) and
`accountId` is added (tenancy).
- `loadActiveRunForContact`, `isDuplicateInbound`,
`findEntryFlow` all query by account_id.
- `startNewRun` inserts into flow_runs with account_id.
- All `engineSend*` calls pass `accountId: run.account_id`.
- `src/lib/automations/meta-send.ts` + `src/lib/flows/meta-send.ts`
- All four engine senders (text, template, media, interactive
buttons/list) accept `accountId` and look up the contact +
whatsapp_config by account_id. user_id stays for inserts.
- `src/lib/flows/types.ts`
- `FlowRow`, `FlowRunRow`, `DispatchInboundInput` gain
`account_id` / `accountId`.
- `src/types/index.ts`
- `Automation` gains `account_id`. Doc note on the user_id
field clarifying it's audit-only post-multi-user.
- `src/app/api/automations/cron/route.ts`
- Pass account_id from the pending row to resumePendingExecution.
- `src/app/api/automations/engine/route.ts`
- Uses `getCurrentAccount()` from #169 to resolve the caller's
account, then dispatches with accountId.
- `CHANGELOG.md` — Fixed entry explaining the inbox / NOT NULL fix.
Backwards compatibility
- All affected tables already have account_id from migration 017.
This PR populates it on insert paths that hadn't been updated.
- The `user_id` columns are kept everywhere — they still identify
the agent / author for audit display. They just no longer
enforce tenancy.
Verification
- `npm run lint` → 0 errors (same 19 pre-existing warnings)
- `npm run typecheck` → clean (had 4 errors mid-refactor; all
resolved by the call-site updates)
- `npm test` → 366/366 pass (engine tests are pure unit
tests over helper functions; not affected)
- `npm run build` → succeeds
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PR 4 of the multi-user accounts series. Lands the link-only
invitation flow — admin generates a shareable URL, copies it,
sends it via any channel. No email service needed.
API surface
- GET /api/account/invitations — list pending (admin+)
- POST /api/account/invitations — create (admin+)
- DELETE /api/account/invitations/[id] — revoke (admin+)
- GET /api/invitations/[token]/peek — public, rate-limited per-IP
- POST /api/invitations/[token]/redeem — authenticated, rate-limited per-IP
Token lifecycle
- POST returns the plaintext token + share URL exactly ONCE in
the response. We store only the SHA-256 hash. There is no
endpoint that resurfaces it; if the admin dismisses the dialog
without copying, the only recourse is to revoke and re-issue.
- Peek and redeem take the plaintext token in the URL path, hash
it in TS via `hashInviteToken`, and look up by `token_hash`.
Two new SECURITY DEFINER RPCs back the cross-RLS work
(migration 019):
- `peek_invitation` — anonymous read. Returns a uniform JSON
envelope: `{ ok, account_name, role, expires_at }` or
`{ ok: false, reason: not_found|used|expired }`. The route
forwards verbatim so the join page can branch cleanly.
- `redeem_invitation` — authenticated. Locks the invite row
`FOR UPDATE`, validates expiry/usage, refuses if the caller is
already in a shared account OR their personal account has any
domain rows (checked across all 11 user-creatable parent
tables), then atomically moves profile.account_id +
account_role, stamps the invitation accepted, and deletes the
orphan personal account. Single transaction.
Error contract
- 22023 (invalid_parameter_value) → 400 — not_found / used / expired
- 42501 (insufficient_privilege) → 401 — caller not authenticated
- 23505 (unique_violation, reused) → 409 — caller's account has data,
or they're already a
member of the inviter's
account / another shared
account
Rate limiting
- Two new buckets in `RATE_LIMITS`: invitationPeek (30/min/IP)
and invitationRedeem (10/min/IP). Belt-and-braces against
token enumeration even though 256-bit entropy makes it
theoretical. Uses the existing fixed-window in-memory limiter.
Why no peek-then-redeem two-step on the server
- Peek is read-only and idempotent; redeem locks `FOR UPDATE` so
two concurrent redemptions can't both win. The peek result is
cosmetic ("show me what I'm joining"); the actual write
happens entirely inside the redeem RPC's transaction, so a
TOCTOU between peek and redeem just means the user sees the
same `{ ok: false, reason }` on the next render.
Verification
- `npm run lint` → 0 errors (same 20 pre-existing warnings)
- `npm run typecheck` → clean
- `npm test` → 366/366 pass
- `npm run build` → succeeds; the 4 new routes register as ƒ
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PR 2 of the multi-user accounts series. Pure helpers and types
that the upcoming API routes (account, members, invitations,
redeem) all depend on. No new API surface yet — this commit is
intentionally a self-contained library + tests so the next PR
can focus on routes without re-litigating the foundations.
What landed
- `src/lib/auth/roles.ts` — `AccountRole` type mirroring the
Postgres enum, `roleRank` / `hasMinRole`, plus the capability
predicates (`canManageMembers`, `canEditSettings`,
`canSendMessages`, `canViewOnly`, `canDeleteAccount`,
`canTransferOwnership`). Pure functions; no I/O.
- `src/lib/auth/invitations.ts` — token generation
(`crypto.randomBytes(32).toString('base64url')`), SHA-256
hashing for at-rest storage, `inviteUrl` builder, and
`inviteExpiresAt` / `clampExpiryDays` for the redeem deadline.
Pure; no Supabase.
- `src/lib/auth/account.ts` — server-only. `getCurrentAccount()`
resolves caller's user + profile + account in one Supabase
round trip; `requireRole(min)` wraps it with a minimum-role
check. Custom `UnauthorizedError` / `ForbiddenError` classes
and a `toErrorResponse(err)` helper that maps them to
`NextResponse` 401/403 with a generic 500 fallback (no
uncategorised error leakage on the wire).
- `src/types/index.ts` — extends `Profile` with `account_id` +
`account_role`; adds `Account`, `AccountMember`, and
`AccountInvitation` interfaces matching migration 017's
shapes. The legacy `profile.role TEXT` column is annotated
for future cleanup.
Tests
- `roles.test.ts` — 16-row hierarchy matrix + every predicate's
boundaries. Includes a "matches the SQL helper's numeric
mapping" check so the JS rank and Postgres `is_account_member`
CASE can't silently drift.
- `invitations.test.ts` — token length / charset / uniqueness,
hash determinism, a known SHA-256 vector (regression net
against the hash function ever changing and orphaning every
stored token_hash), URL formation edge cases, and the
`clampExpiryDays` envelope (default / non-finite / negative /
oversize).
Verification
- `npm run lint` → 0 errors (same 20 pre-existing warnings).
- `npm run typecheck` → clean.
- `npm test` → 366/366 pass (319 prior + 47 new).
- `npm run build` → succeeds.
What's NOT in this PR
- No API routes yet — `/api/account/*` and `/api/invitations/*`
land in PR 3 + 4 of the series.
- No UI changes — useAuth extension, Members tab, and Join page
come later (PRs 5-7).
- No WhatsApp webhook account routing (PR 8).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Closes the editing-affordance gap from PR 2a. The canvas is now in
full parity with the list view for everything that fits in a node
graph: connections, deletes, and adding nodes all work on-canvas.
Adds:
- Per-slot Handle components on the custom node renderer. Single-
outgoing types get one handle on the right edge; condition gets
`true` / `false` handles; send_buttons and send_list render slot
rows inside the card with one handle each, so the user can SEE
which button or list row they're wiring. Every non-start node
gets a target handle on the left.
- Drag-to-connect (`onConnect`). Looks up the source node, dispatches
the right config patch via the new `applyEdgeConnection(node,
sourceHandle, target)` helper, and routes through updateNodeConfig.
Rejects self-loops silently (a button whose target is its own node
= infinite reprompt).
- Keyboard delete on nodes (Backspace + Delete, both for cross-OS
parity). Routes through removeNode, which now ALSO unlinks every
inbound reference via the new `unlinkNodeReferences(nodes, key)`
helper. No more dangling arrows after a delete — the validator
previously flagged these as errors and the user had to chase them
down manually.
- Edge delete. Selecting an edge + Delete clears just that slot's
next_node_key (rather than removing anything), via the same
applyEdgeConnection helper passing "" as the target.
- Floating "+ Add node" button (React-Flow Panel, bottom-right).
Same dropdown as the list view; drops the new node at the visible
viewport center via reactFlow.screenToFlowPosition.
New pure helpers in src/lib/flows/edges.ts (no React, fully unit-
tested — 18 new tests, 319 total passing):
- outgoingSlots(node) — list of {id, label} per outgoing slot.
- applyEdgeConnection(node, sourceHandle, targetKey) — config patch
for a drag-to-connect, or null for unrecognised handles.
- unlinkNodeReferences(nodes, deletedKey) — clears every direct +
nested reference to deletedKey. Returns nodes-by-identity when
no reference matches to avoid needless re-renders.
The hook's removeNode is the only consumer of unlinkNodeReferences
right now, which means LIST-view deletes also stop leaving dangling
refs — strict improvement for both views.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The second half of the user feedback that drove send_media: the
linear-list editor reads as a checklist, not a branching flow, which
made conversational structures hard to understand at a glance. This
introduces a canvas view alongside the list — boxes + arrows, drag-
zoom-pan, labeled edges per branch slot. Canvas is the new default;
the list stays as a fallback toggle (per-browser preference,
persisted to localStorage).
PR 1 scope is intentionally read-only: render, don't edit. The
foundation is testable in isolation before adding drag-to-persist,
side-panel editing, and drag-to-connect (PR 2) and polish like
keyboard shortcuts, mobile fallback, and validator integration
(PR 3).
Pieces:
- New @xyflow/react (~30 KB gz) + @dagrejs/dagre (auto-layout for
pre-existing flows whose position_x / position_y are all zero —
flow_nodes already reserved those columns in migration 010
precisely for this).
- Shared module src/components/flows/shared.tsx lifts NodeType,
BuilderNode, NODE_META, summarizeNode, truncate so both views
render nodes identically.
- Pure helpers (no React, fully unit-tested):
src/lib/flows/edges.ts — derive labeled canvas edges per
outgoing slot (button, list row,
condition branch)
src/lib/flows/layout.ts — dagre wrapper + shouldAutoLayout
guard so manually-positioned flows
never get rearranged on open
- FlowCanvas — client component, custom node renderer matching the
list view's collapsed-card styling, MiniMap + Controls, attribution
hidden.
- FlowEditorShell — owns the Canvas / List toggle, conditionally
mounts the right view.
21 new unit tests (layout + edges); existing 271 still pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Lets flows attach an image, video, or document (e.g. PDF invoice /
receipt) to a customer message — the most-requested gap in v1 was that
nodes were text-only. Single node with a `media_type` discriminator
rather than three node types: Meta's wire payload differs only in the
top-level key, so three types would triple the surface area (forms,
engine cases, menu entries) for no behavioural difference.
Files are uploaded directly from the builder to a new public
`flow-media` Supabase Storage bucket (per-user RLS, 16 MB cap, MIME
allowlist) and the resulting public URL is what Meta fetches at send
time. Same trade-off as the existing avatars bucket.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Closes the silent-failure bug a user reported: when multiple phone
numbers are connected to the same WABA, only the first one actually
receives webhook events from Meta. The wacrm UI shows "Connected"
because the metadata probe succeeds, but Meta's activity log is
empty for every subsequent number.
Root cause
----------
The POST /api/whatsapp/config handler verified credentials with a
GET /{phone_number_id} probe and saved them to Supabase — but
never called Meta's two registration endpoints that actually wire
a number for inbound events:
1. POST /{phone_number_id}/register
Subscribes the number to THIS app's webhook. Requires a
6-digit 2FA PIN the user previously set in Meta WhatsApp
Manager → Two-step verification. Without it, Meta routes
events for that number to whichever app last claimed it.
2. POST /{waba_id}/subscribed_apps
Subscribes the WABA itself to this app. Required once per
WABA, but idempotent on Meta's side so we call on every
save without harm.
The first number "worked" because someone (the user, Embedded
Signup, or Meta dashboard) had already done both calls for it
out-of-band. Every subsequent number silently failed.
What this PR does
-----------------
* New helpers in meta-api.ts: registerPhoneNumber (treats Meta's
"already registered" 400 as success), subscribeWabaToApp
(idempotent), getSubscribedApps (for the diagnostic).
* POST /api/whatsapp/config now:
- Accepts an optional `pin` (6-digit, validated).
- Calls subscribeWabaToApp on every save (idempotent).
- Calls registerPhoneNumber on first save and whenever a PIN
is supplied. Skipped when the same number is already
registered locally and no fresh PIN is provided — re-running
/register with a stale PIN would actually undo the active
subscription.
- Persists registered_at / subscribed_apps_at timestamps and
last_registration_error for diagnostic surfacing.
- Returns {success, registered, registration_error} so the UI
can distinguish "creds valid" from "actually live".
- Saves the row even when /register fails so the user can
retry without re-entering everything.
* New diagnostic GET /api/whatsapp/config/verify-registration —
hits Meta to confirm the saved number is reachable + the WABA
is subscribed. Returns a per-check breakdown the UI renders.
Solves the misleading-green-banner UX failure that caused the
bug to go unnoticed in the first place.
* Migration 015 adds registered_at, subscribed_apps_at,
last_registration_error to whatsapp_config. All nullable so
existing rows survive — the UI labels them "registration status
unknown" until the user clicks Verify or re-saves.
* whatsapp-config.tsx:
- New PIN input with help text pointing to the exact Meta
Manager setting and explaining why it's needed.
- Two-tier status: existing "Credentials valid" alert (relabeled
from "Connected" since it was misleadingly broad) PLUS a new
Registration Status alert that flips emerald/amber based on
the registered_at timestamp.
- "Verify with Meta" button on the registration alert pings
the diagnostic endpoint and renders the per-check result
inline. Surfaces last_registration_error verbatim with a
pointer to retry.
- PIN clears from state after a successful save so it can't
accidentally retrigger /register on the next save.
Tests + checks
--------------
* 9 new fetch-mock tests cover registerPhoneNumber (happy path,
already-registered passthrough, PIN-required error verbatim,
generic Meta error), subscribeWabaToApp (URL + bearer + error),
getSubscribedApps (data, empty, error).
* 265/265 tests pass. Lint 0 errors. Typecheck clean. Build
registers both /api/whatsapp/config and
/api/whatsapp/config/verify-registration.
Out of scope
------------
* Multi-number-per-CRM-user (current schema is UNIQUE(user_id)
so one CRM user = one number). The reported bug was
multi-number-per-WABA across multiple CRM users, which this
fully resolves.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
A focused cleanup PR that addresses the high-signal findings from
a self-review of PRs #148-153. Every change is local to the
templates surface; no schema or wire-format changes.
Safety
------
* Delete now goes through a confirmation Dialog (mirrors the
existing tag-manager pattern). Copy explicitly tells the user
when Meta will be hit, so a slip on the trash icon doesn't take
the template off Meta as well as locally.
* AUTHENTICATION submit is now blocked by a persistent in-dialog
banner + disabled submit button. The disappearing toast that
used to fire on submit is gone — defensive `return` remains in
handleSubmit as a second line of defense.
* updateButton uses a per-variant switch + ButtonPatch union
instead of the partial-spread + `as TemplateButton` cast. Each
variant's spread keeps its discriminant pinned; TS enforces
exhaustiveness.
* New template-row-guard.ts surfaces malformed Supabase rows at
the route boundary instead of letting buildSendComponents crash
deep in the stack with an opaque TypeError.
* UUID regex check on /[id] PATCH+DELETE so non-UUID payloads
return a clean 400 before hitting Supabase.
Submit-route cleanup
--------------------
* Two near-identical upsert blocks (error path + success path)
collapsed into a buildUpsertRow + upsertTemplateRow helper.
Adding a column now touches one place, not two.
UX copy + feedback
------------------
* Submit success toast sets the right expectation: "typical
review time is 24 hours. Status updates automatically." Old
copy just said "status will update once approved."
* Header media URL helper text is now actually useful: HTTPS
requirement, ~24h lifetime, per-format size/format recommendations
(JPEG/PNG ≥800×418 ≤5 MB for images, MP4 ≤16 MB for video, etc).
* Truncated-sync toast switches from `warning` (which auto-dismisses
fast) to `error` with `duration: 10000` and concrete copy: "Synced
the first 2000 templates only — your account has more. Sync again
to continue."
* Edit and Resubmit buttons gain visible text labels alongside
their icons + better aria-label/title pairs. A lone rotate icon
wasn't self-explanatory.
* Broadcast picker drops the redundant "Approved" chip — every
template in that list is APPROVED by filter, so the chip carried
zero information. Unused templateStatusConfig import removed.
State + a11y
------------
* handleSubmit now awaits fetchTemplates BEFORE closing the dialog
so re-opening immediately doesn't see a stale list.
* Header content / media URL / sample are preserved across format
switches in the New Template form. The payload builder is already
format-aware, so orphan values on hidden fields are harmless —
and keeping them lets users switch formats to compare without
losing typing.
* Button-type Select gains the same null-guard the header Select
got in PR #148 (per @base-ui semantics, onValueChange can fire
with null on deselect).
* Dynamic body-sample inputs get id + aria-label so each one is
individually addressable (WCAG A — screen-reader announcement
beat "edit text"). Header text + header sample inputs too.
Deliberately NOT here
---------------------
* The 940-line template-manager component is not split. It's
cohesive (one feature surface) and splitting for size alone
would be over-engineering.
* extractVariableIndices in the buttons loop is not memoized.
Regex on a short URL string is sub-microsecond; a real profiler
would never flag it.
* No JS heap secret-clearing for access tokens. Strings are
immutable in JS — there's no way to actually clear one.
Tests: 256/256 pass (no new tests required — pure refactor +
copy + new UI elements). Lint 0 errors. Typecheck clean. Build
succeeds.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Closes the last hole left by PR 3 — until now, sendTemplateMessage
only emitted a body component, so templates with image/video/
document headers or URL-with-{{1}} buttons failed at send time
with Meta's "missing component" 400. After this PR, every
APPROVED template that wacrm can submit is also fully sendable.
New helper
----------
src/lib/whatsapp/template-send-builder.ts builds the per-send
Meta components array from a MessageTemplate row + caller-supplied
SendTimeParams. It's a different beast from template-components.ts
(which builds the SUBMISSION shape) because the send-time payload
has a different schema: header type-specific parameter objects,
button components with sub_type + index, body_text without a 2D
example wrapper, etc.
Auto-fills as much as possible from the template:
* static IMAGE/VIDEO/DOCUMENT headers ride along using the row's
header_media_url (or header_handle when present — preferred)
* URL buttons without {{1}} need no per-send payload
* QUICK_REPLY / PHONE_NUMBER buttons need nothing
* COPY_CODE buttons fall back to the template's example so the
most common "static promo code" case Just Works without UI
Variable-bearing fields require per-send values:
* TEXT header with {{1}} → messageParams.headerText
* Body variables → messageParams.body[]
* URL button with {{1}} → messageParams.buttonParams[i]
* COPY_CODE override → messageParams.buttonParams[i]
19 unit tests assert the wire shape per scenario — header types,
URL button at correct index when QR buttons precede it, example
fallback, and the throw-with-specific-message paths.
sendTemplateMessage extension
-----------------------------
src/lib/whatsapp/meta-api.ts gains optional `template` +
`messageParams` args. When `template` is provided, the full
components array is built via the new helper. Legacy `params:
string[]` (body only) path is preserved as-is — old callers
behave identically. When both shapes are present, structured
messageParams.body wins.
Route + UI wiring
-----------------
POST /api/whatsapp/send and /api/whatsapp/broadcast now load the
template row by (user_id, name, language) and pass it through.
Broadcast loads once per request (not per recipient — avoids N+1).
Both accept `template_message_params` / per-recipient
`messageParams` from the client.
src/components/inbox/template-picker.tsx scans the selected
template for variable-bearing fields and renders an input per
need: header text variable, body variables (existing), URL
button suffixes. Per-button preview shows the final URL with
the substitution so the user sees what the recipient will see.
The callback shape changes to `(template, { body, headerText?,
buttonParams? })`; message-thread.tsx updated accordingly.
Out of scope
------------
Broadcast composer doesn't yet collect per-recipient header /
button params — the legacy body-only flow continues to work for
broadcasts. Templates with header / URL-button variables broadcast
fine if every recipient gets the same value via API call shape;
the UI for varying it per recipient lands in a follow-up.
This is the last PR of the Issue #147 overhaul sequence.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Wires Meta's three template-lifecycle webhook events into the
existing /api/whatsapp/webhook route so APPROVED / REJECTED /
PAUSED transitions land locally without a manual Sync click.
Handled events
--------------
message_template_status_update
→ flip status to the raw Meta enum (re-using normalizeStatus
from PR 3 so PENDING_REVIEW collapses to PENDING etc.)
→ persist `reason` to rejection_reason on REJECTED so the
red banner in template-manager shows what to fix
→ clear rejection_reason + submission_error on any other
transition so a stale REJECTED message doesn't linger
after re-approval
message_template_quality_update
→ set quality_score (GREEN/YELLOW/RED). Unknown values land
as NULL so the chip just disappears.
message_template_components_update
→ info-log only for v1. Auto-persisting Meta's modified
components would silently change what the user thought
they submitted; instead we point them to "Sync from Meta"
in the log line. A follow-up PR could surface a banner.
Lookup by meta_template_id (indexed in migration 014) — globally
unique per WABA, so no user_id filter needed. Unknown rows log a
warning instead of failing; >1 matches log a warning too in case
of cross-tenant drift.
New module
----------
src/lib/whatsapp/template-webhook.ts isolates the dispatch + each
handler. The route handler just calls
`handleTemplateWebhookChange(change, supabaseAdmin())` when
`isTemplateWebhookField(change.field)` matches, and the existing
messaging branches are short-circuited so a template event doesn't
try to read message-shaped fields off the value.
12 new tests cover: status APPROVED happy path, REJECTED with
reason, REJECTED missing reason → generic fallback, PENDING_REVIEW
normalization, missing meta_template_id no-op, unknown row warning,
quality score update, invalid quality value → NULL, components
update no-op, unknown field defensive no-op, and the field detector.
Setup requirement
-----------------
Subscribing to these fields is a one-time manual step in Meta App
Dashboard → WhatsApp → Configuration → Webhooks (no API for Cloud
API apps). Documented in a top-of-file comment block in
template-webhook.ts. Until it's done, status updates still arrive
via the manual "Sync from Meta" button — that fallback stays.
Out of scope (final PR in sequence):
* sendTemplateMessage media + button param support (PR 6)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds the missing lifecycle operations so users can manage existing
templates without leaving wacrm.
API surface
-----------
PATCH /api/whatsapp/templates/[id]
Edits (and re-submits) an existing Meta-side template. Used by
the "Edit" action on APPROVED rows and "Resubmit" on REJECTED /
PAUSED rows — same operation, different UI labels. Meta replaces
components wholesale and flips status back to PENDING for
re-review. Status guard rejects edits on PENDING / DISABLED /
IN_APPEAL / PENDING_DELETION and on local-only (no
meta_template_id) rows. Honors WHATSAPP_TEMPLATES_DRY_RUN.
DELETE /api/whatsapp/templates/[id]
Calls Meta `DELETE /{waba_id}/message_templates?name=X&hsm_id=Y`
when meta_template_id is set — hsm_id is critical, without it
Meta deletes EVERY language variant of the same name. Treats
Meta 404 as a no-op (template already gone). Always removes the
local row.
New Meta helpers
----------------
src/lib/whatsapp/meta-api.ts gains editMessageTemplate and
deleteMessageTemplate following the established named-args +
throwMetaError pattern. Eleven new fetch-mock tests assert the
exact wire shape — especially that hsm_id always rides along when
provided, which is the difference between "delete one variant"
and "delete every variant."
UI changes
----------
src/components/settings/template-manager.tsx:
* Edit (pencil) button visible for APPROVED rows
* Resubmit (rotate) button visible for REJECTED / PAUSED rows
— same flow as Edit, different label
* Delete now calls the route handler (so Meta gets the call too)
with a spinner during the in-flight request
* Dialog pre-fills form from the row when opened in edit mode
* Dialog title / CTA / description switch between "New Message
Template" + "Submit for Approval" and "Edit Message Template"
+ "Save & Resubmit"
* Name + language inputs are disabled in edit mode (Meta keys
templates by id; changing either is effectively a new template)
* Closing the dialog clears editingId so the next open starts
fresh
Design decision
---------------
One PATCH endpoint instead of the two endpoints the plan called
for (PATCH for edit, POST for resubmit). The Meta call is the same
operation either way — the difference is UI semantics, handled in
the component. Cuts duplication; easy to split later if needed.
Out of scope (final PR remaining in the sequence):
* Webhook handler for real-time status (PR 5)
* sendTemplateMessage media + button param support (PR 6)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Closes the second half of #147. Users can now create a template in
wacrm and submit it to Meta for approval without leaving the app —
no more dummy local rows that fail with Meta error #132001 on send.
API surface
-----------
POST /api/whatsapp/templates/submit
Auth → fetch whatsapp_config → validate payload → build Meta
components → POST to /{waba_id}/message_templates → persist row
with meta_template_id + status + sample_values +
last_submitted_at. Per-field validators throw with specific
messages before the network call, so misconfigured templates
fail with "button #3 missing url" instead of a 400 from Meta.
429s are detected and surfaced as "rate limit hit (100/hr)".
WHATSAPP_TEMPLATES_DRY_RUN=true short-circuits the Meta call and
issues a synthetic `dry-run-<uuid>` meta_template_id so CI and
local dev can exercise the full form without a real WABA.
New modules
-----------
src/lib/whatsapp/template-validators.ts
Pure functions encoding Meta's caps: 1024-char body, 60-char
footer/header, 25-char button text, 10 buttons max, 2 URL / 1
PHONE_NUMBER / 1 COPY_CODE, QUICK_REPLY contiguity rule, URL
example required when {{1}} is present, COPY_CODE example
required, header media URL must be http(s), variables must be
contiguous starting at {{1}}. 41 unit tests cover each rule.
src/lib/whatsapp/template-components.ts
Translates our local row shape into Meta's components array
(HEADER → BODY → FOOTER → BUTTONS, body_text examples as a 2D
array per spec, header_handle preferred over header_url when
both are present). Snapshot tests assert the wire shape.
src/lib/whatsapp/template-status-normalize.ts
Single source of truth for upstream→local status mapping.
Replaces the duplicated normalizer in sync/route.ts; the
upcoming webhook handler will share it too.
src/lib/whatsapp/meta-api.ts
New submitMessageTemplate() helper following the existing
named-args + throwMetaError pattern.
Form rewrite
------------
src/components/settings/template-manager.tsx replaces the flat
form with a component-based builder:
* Header: format select (None / Text / Image / Video / Document)
+ conditional inputs — text + variable-aware sample row for
TEXT, URL paste for media (file upload slot ready for the
Resumable Upload follow-up)
* Body: textarea with live {{N}} extraction → sample-value rows
auto-appear and shrink as variables are added/removed
* Footer: 60-char capped input
* Buttons editor: add up to 10, per-row type picker
(Quick Reply / URL / Phone / Copy Code), conditional URL +
example + phone + code fields, X to remove
* Card display now surfaces rejection_reason / submission_error
(red banner) and quality_score (GREEN/YELLOW/RED chip)
* AUTHENTICATION shows a "create in Meta Manager + Sync" banner
and disables Submit — that category has a different shape
(fixed body, mandatory OTP button) and lands in a follow-up
Side effects
------------
* Broadcast Step 1 picker (step1-choose-template.tsx) now
filters to status='APPROVED' to match the inbox picker —
selecting a non-approved template would 400 on send.
* .env.local.example documents META_APP_ID (Resumable Upload
follow-up) and WHATSAPP_TEMPLATES_DRY_RUN.
Out of scope (subsequent PRs):
* Edit / resubmit / delete endpoints (PR 4)
* Webhook handler for real-time status (PR 5)
* sendTemplateMessage media + button param support (PR 6)
* Resumable Upload + AUTHENTICATION builder (follow-up)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Switches message_templates.status from the local TitleCase set
(Draft|Pending|Approved|Rejected) to Meta's raw enum (DRAFT|PENDING|
APPROVED|REJECTED|PAUSED|DISABLED|IN_APPEAL|PENDING_DELETION) and
adds the columns the upcoming submit / edit / resubmit / webhook
flows need. No new user-facing behavior — this is groundwork.
Migration 014 also adds:
* sample_values JSONB Meta requires these for variable templates
* meta_template_id TEXT hsm_id for scoped edit/delete (avoids
nuking sibling-language variants)
* rejection_reason TEXT surfaced from webhook on REJECTED
* quality_score TEXT GREEN | YELLOW | RED, from webhook
* header_handle TEXT from Resumable Upload (media headers)
* header_media_url TEXT URL fallback for media headers (v1 path)
* submission_error TEXT last 4xx from Meta, for retry surface
* last_submitted_at rate-limit awareness (100 creates/hour)
* buttons JSONB shape CHECK permissive array-of-typed-objects guard
(strict per-type validation lives in API)
* UNIQUE (user_id, name, language)
so sync upsert is atomic and prevents
duplicate language variants
* idx on meta_template_id fast webhook lookup
Sync route now:
* persists Meta's status verbatim (PAUSED no longer collapses to
REJECTED — these are different states with different recoverability)
* parses BUTTONS components into the new discriminated-union shape
* extracts sample_values from body.example.body_text + header.example
* persists meta_template_id, quality_score, header_handle
Shared template-status.ts gives the manager / inbox picker / broadcast
picker one source of truth for labels + dark-theme badge classes.
Part of the Issue #147 overhaul plan (PR 2 of 6 in the sequence).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The Flows feature was previously gated on a per-account
`profiles.beta_features` opt-in: the sidebar entry was invisible
to non-beta accounts, every /api/flows route 404'd, the webhook
dispatcher was a no-op, and the agent-handoff pause was skipped.
This PR opens the surface to every authenticated user. A small
"Beta" chip in the sidebar nav + on the /flows page H1 keeps the
"young feature" signal without holding anyone out.
Removed:
- `src/lib/flows/feature-flag.ts` + its tests (-11 tests).
- The `requireFlowsBeta` / `requireFlowsBetaAndOwnership` helpers
in the API routes — collapsed to plain `requireUser` / RLS
ownership.
- The `isProfileFlowsEnabled` helper + branch in the webhook
dispatcher — the runner takes the cheap "no_match" exit on
accounts with no active flows (single indexed SELECT).
- The agent-handoff pause gate in /api/whatsapp/send — the UPDATE
matches zero rows when no active run exists, also cheap.
- Client-side beta gates in the 3 flows pages (and the redundant
`useAuth` + `profileLoading` plumbing they needed).
- The defensive message-bubble gate that would have rendered
interactive replies as plain text for non-beta accounts.
- The `betaKey` field on the sidebar NavItem schema, replaced by a
simpler boolean `beta` that purely drives the visual chip.
Kept:
- `profiles.beta_features` column. Generic mechanism for future
beta gates; no current feature reads it. Comments in
`src/types/index.ts` + `use-auth.tsx` updated to say so.
- The chip can be removed in a one-liner once the team's satisfied
Flows has bedded in.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PR #132 (color themes) only tokenized the high-visibility chrome.
Inbox / contacts / pipelines / broadcasts / automations / auth /
settings sub-forms still had ~150 hard-coded `violet-*` classes
across ~32 files, so picking a non-violet theme felt half-broken —
sidebar and header changed colour, but message bubbles, deal cards,
and the broadcast wizard all stayed violet.
This is the mechanical follow-up: replace every `violet-N(/N)?`
with `primary(/N)?` so the theme picker themes the whole app.
Two non-trivial swaps inside the pass:
- `text-white` on a `bg-primary` button → `text-primary-foreground`.
Emerald and Amber define a *dark* `--primary-foreground` because
their primaries are bright; `text-white` over those would tank
contrast.
- `bg-primary hover:bg-primary` (hover identical to base — original
code had this latent bug from `hover:bg-violet-500`) → adds `/90`
on hover so the lift is visible.
Test update:
- `broadcast-status.test.ts` asserted shade numbers (`bg-X-N/10`).
Relaxed the regex to also accept token form (`bg-X/10`) since
brand-accent status badges now ride the active theme.
Files touched: 49. Lint clean (existing warnings only), 173/173
tests, typecheck + build green.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds a new "Appearance" tab in Settings with 5 dark color themes
(Violet, Emerald, Cobalt, Amber, Rose). Pick a theme → applied
immediately and persisted to localStorage. Boot script in the root
layout replays the pick before first paint so there's no
flash-of-default on subsequent loads.
Implementation, ordered the way it's read:
1. `globals.css` — 5 `html[data-theme="..."]` blocks extracted
verbatim from the design handoff at project/Color Themes.html.
`:root` mirrors Violet so SSR / first-paint renders correctly
even before the boot script runs. Three new tokens
(`--primary-hover`, `--primary-soft`, `--primary-soft-2`) added
to all themes — used by tinted-primary surfaces.
2. `src/lib/themes.ts` — single source of truth for the catalog
(id, name, tagline, swatch). Adding a theme = append a CSS
block + append an entry here.
3. `src/hooks/use-theme.tsx` — ThemeProvider client component. Reads
what the boot script already applied, syncs across tabs via the
`storage` event, persists to localStorage. Falls back to a no-op
setter when used outside the provider.
4. `src/app/layout.tsx` — inline `beforeInteractive` script that
reads localStorage and sets `documentElement.dataset.theme`
before React hydrates. Knowledge of valid theme IDs comes from
the THEME_IDS constant so adding a theme doesn't silently break
the boot path. Wraps the tree in ThemeProvider.
5. `src/components/settings/appearance-panel.tsx` — picker UI. 5
cards, swatch + name + tagline + a 4-tone preview strip. Active
card carries a primary-tinted ring + "Active" chip.
6. `src/app/(dashboard)/settings/page.tsx` — new "appearance" tab
alongside profile/whatsapp/templates/tags.
7. Key-surface tokenization — replaced hard-coded `violet-*` with
`primary` tokens on the high-visibility surfaces so the theme
switch feels real: sidebar logo + active nav + unread dot +
avatar fallback, header avatar, dashboard-shell loading
spinner, Flows header icon + entry-node picker icon + entry
badge + flash ring, Flows list page (template card hover,
FlowCard icon), Flows editor + runs "back" links, Settings tabs
active state.
NOT touched (intentional follow-ups):
- Per-node-type colors in flow-builder.tsx NODE_META — those are a
deliberate per-type rainbow, not a brand accent.
- The inbox/contacts/pipelines surfaces — many `violet-*` usages
there. Sweep is a contained follow-up; the user explicitly
scoped this PR to "tokens + tab + key surfaces".
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
