PR #10 added 000035_api_key_suffix and PR #11 separately added
000035_api_key_smart_limits, both targeting the api_keys table.
golang-migrate refuses to load when two source files share a
version number, so the backend container failed health checks
in fresh stacks.
Bumping api_key_smart_limits to 000037 keeps PR #10 in its
original slot (it landed first) and matches the next free
version after 000036_contact_categories. The two migrations
touch different columns so apply order does not matter.
- gofmt-align ContactEngagement fields and drop trailing blank line
in contact/export.go
- remove no-op self-assignment ac.CustomFields = ac.CustomFields
flagged by govet
Disable the linters that fire on legacy code without flagging real
bugs: `unused` (orphan repos kept for future feature flags),
`unconvert` (defensive type conversions), `gosimple` (style
suggestions in code we don't want to touch).
govet: disable `shadow` (idiomatic `err :=` re-decls in transaction
patterns) and `nilness` (legitimate defensive nil checks that look
tautological to the analyzer).
Ran `gofmt -w internal/ cmd/` — every Go file now passes
gofmt -l with no output.
Kept: govet, staticcheck, ineffassign, typecheck, bodyclose, noctx,
sqlclosecheck, gofmt, goimports, misspell — the real-bug checks.
Backend:
- Fix contact-create 500 (nil custom_fields, doubled slice, bad RETURNING SQL)
- Avatar upload: migration 000033, S3 public-read, PNG/JPG only,
client-resized to 512px + server dimension cap (1024px)
- Pull avatar_url through user + organization repo queries
Frontend:
- Settings restructured into nested routes with a rail layout
(/app/settings/{profile,notifications,security,members,roles,
workspace,billing,danger}); flat Section/Row primitives replace
the per-card rectangles; Save buttons only render when dirty
- Standalone /app/billing and /app/team removed; legacy URLs
redirect to the settings sections; UserNav trimmed accordingly
- CRM rebuilt: Pipelines CRUD + stage editor, Deals kanban with
HTML5 drag/drop, Tasks bucketed by due-date with inline toggle.
Frontend models realigned with backend (Deal.name, CRMTask.status
enum, paginated list shapes)
- Avatars: AvatarUploader component, client-side canvas resize,
wired into Profile + Workspace settings; UserNav + OrgSwitcher
render the uploaded image with initials fallback
- RBAC: lib/permissions.ts mirrors organization_permission.go;
inline role picker in Members; Roles & access section shows the
permission matrix and per-role member counts
- Audit log page at /app/audit, gated to owner+admin via canManage
- Plans aligned with warmbly-web pricing: Starter/Grow/Business/
Enterprise via lib/plans.ts; PlanPill, billing page, sidebar
badges and LockedSurface all read from the same catalogue
- Header PlanPill shows current plan with status-aware coloring;
sidebar locked rows show the required-plan badge instead of a
generic lock icon
Perf:
- QueryClient defaults (staleTime: 30s, refetchOnWindowFocus: false,
retry: 1) — kills 3-5 round-trip storm on every navigation
- useSubscription, usePlans → staleTime: Infinity (only invalidate
on plan-change mutations); useUser/Timezones/Orgs get long stales
with refetchOnMount: false
- vite.config: optimizeDeps for heavy libs + server.warmup for the
most-mounted entry pages
Inbox filter:
- Backend: MailSearchParams gained EmailAccountIDs []uuid.UUID; the
search SQL filters with `email_id = ANY($)`. /unibox handler now
accepts both `email_id=<uuid>` (legacy) and `email_ids=<csv>`.
- Frontend: UniboxSearchParams gained accountIds[] and a UI-only
tagId. searchIncoming sends email_ids=csv. UniboxFilterSheet:
Accounts section is now (a) a row of tag chips backed by user.tags
with per-tag account counts and (b) a multi-select list of every
connected mailbox with an inline checkbox + avatar; accounts that
belong to the active tag get a "via tag" affordance. Picking a tag
resolves to the underlying account IDs at Apply time. "Select all"
/ "Clear" inline in the SectionBar header.
Org gate + onboarding:
- New /select-org page. Three sections: pending invitations (one-
click Join), existing memberships (pick one to enter), and a
Create New Workspace form (slate-900 primary). Routed at
/select-org.
- OrgGate hook lives inside RealtimeManager. On load, if the user
has zero orgs and no current org, it navigates to /select-org
replace. Renders null so it doesn't displace AppLayout.
Invite + join:
- Team page rebuilt with real data: useMembers + usePendingInvitations,
plus InviteDialog (email + role popover, slate-900 send button).
Inline remove on member rows (skip "owner"), inline cancel on
pending invitations.
- Pending invitations show up on /select-org too — a freshly
invited user can accept without ever entering the dashboard first.
Response unwrapping:
- Org/member/invitation list clients now tolerate the backend's
{data: T[] | null} envelope (it's the consistent shape across the
Go handlers). Map nested membership rows into the flat
Organization shape the rest of the app expects.
Seeder: re-run verified — dev@warmbly.com still gets "Dev's
Organization" so they don't bounce through /select-org.
User: "inbox is really really bad. So I want all possible ways to
search for an email that we can do... realtime for everything and the
dashboard to show our latency... show how much unread emails." Plus a
follow-up: "I don't like how the emails looks like because they have
that blue gradient, I want dashboard style good one."
Inbox:
- Wired the backend search endpoint (GET /unibox with from/subject/
unseen/since/until/cursor/limit) — was implemented server-side but
the frontend was never calling it. Inbox now actually reflects
server data.
- New UniboxSearchParams model + searchIncoming client + infinite
useUniboxSearch hook that drops null rows defensively.
- ConversationList: SearchInput (subject substring) + quick-filter
strip (All / Unread / Today / This week). Unread count surfaces
in the SectionBar header AND on the Unread chip. Skeleton +
explicit error block with retry; "Load more · N shown" when more
pages are available.
- UniboxFilterSheet (advanced filters in the right-side panel):
free-text query, sender substring, account picker pulled from
the user's connected mailboxes, status toggle (Any/Unread/Read),
since/until date pickers with toggle, newest/oldest sort. Draft
state mirrors parent until Apply.
LivePanel telemetry (sidebar):
- Real WS roundtrip latency. SocketProvider stamps performance.now()
per heartbeat ref; phoenix phx_reply with that ref computes the
delta and publishes via setWsLatencyMs. LivePanel colour-codes
the latency text: <100ms emerald, <300ms amber, ≥300ms red, "—"
when disconnected.
- Unread count row reads from useAppStore.unseenCount.
- Status label: OFFLINE / CONNECTING / LIVE (with pulse) / IDLE,
tied to connectionStatus + active mailbox count.
Transactional emails (no more blue gradient):
- base.go rewritten as dashboard chrome: cream #f5f6f8 background,
white card with hairline #e2e8f0 border, 8px radius, slate-900
text. Logo monogram in slate, no decorative haze, no gradients.
- login_code / registration_code: tiny uppercase eyebrow + 18px
bold heading + neutral body + monospace code pill in a hairline-
bordered box. No serif type.
- reset_password / welcome: same chrome. Slate-900 primary button
replaces the sky-gradient one. Plaintext link below for accessible
fallback.
- Template tests updated against the new markup; all green.
User: "when I click on delete the confirm appears behind the form and
it looks really bad, doesn't fit in the theme; and also after I reload
the page, nothing appears after creation".
Two distinct bugs:
1) Confirm dialog stacking + styling
FoldersModal/TagsModal render at z-[110]. ConfirmProvider rendered
the confirm overlay at z-101 with bg-black/30 + scale animation +
poppins styling — visually it landed BEHIND the folders modal and
clicks went through to the backdrop instead.
Rewrote ConfirmProvider in the brae chrome:
- z-[200] so it stacks above page-level overlays AND nested
dialogs.
- Hairline-bordered card, 48px header (red alert tile + "Confirm"
eyebrow), prose body, slate-900 footer (Cancel / red Confirm).
- Escape closes; backdrop closes (both gated on !loading).
- Spinner inside Confirm during the awaited action.
2) Created folders/tags disappeared after page reload
POST /folders + /tags persisted to Postgres fine. The frontend
optimistic-updated the cached user via setQueryData. But
/auth/me did not return folders/tags/categories — the User payload
omitted them entirely. On reload the cache refetched /auth/me,
got missing fields, defaulted to [], and the items vanished from
the UI.
Backend fix:
- models.User now carries Folders/Tags/Categories ([]Group),
always serialized as arrays.
- GroupRepository + GroupService gained a List(ctx, userID)
method; ordered by position then created_at.
- /auth/me handler now calls List on FolderService, TagService,
CategoryService and attaches them to the user before responding.
Verified end-to-end:
GET /auth/me → 200 with full folders/tags arrays populated.
Create a folder, reload the page → folder still in the list.
Two real bugs surfaced from "All folders / Newest dropdowns don't open"
and "hex color must be a valid string":
1) Dropdowns silently no-op (broken across the whole dashboard)
PopoverMenuTrigger asChild uses React.cloneElement to inject
onClick / ref / aria-expanded onto the trigger child. SelectButton
was a plain function component that destructured a fixed prop set
and rendered its own <button> — so the injected props were
dropped on the floor. Click did nothing.
Fix: SelectButton is now React.forwardRef + spreads {...rest} onto
the inner button. The injected click handler reaches the real
element, the dropdown opens, the menu renders, and selection
actually applies state.
Every PopoverMenu trigger using SelectButton was affected — that's
campaigns (folders + sort), emails (tag filter), contacts (sort +
filters page rows). All now work.
2) Adding a folder/tag failed with "hex color must be a valid string"
The /folders + /tags POST landed on groupRepository.Create with
an empty color and the validator rejected. Even before the color
check, the INSERT used tx.QueryRow + Scan against an INSERT with
no RETURNING clause, which always errored with
"sql: no rows in result set" once it got past validation.
API improvements (kept the design but made it forgiving):
- Color defaults: if the request omits color, the server picks one
from an 8-swatch palette based on the new item's position. Two
consecutive creates won't end up identical. Non-empty but
invalid still 400s — that's a client bug worth surfacing.
- Title min length 3 → 1. "Q1", "VIP", short names are common
and shouldn't fail. Trimmed before validation so " " doesn't
pass.
- INSERT now uses tx.Exec instead of QueryRow.Scan — the broken
code would never reach success even when validation passed.
Verified end-to-end:
POST /folders {"title":"Q1"} → 200, color=#94a3b8 (default).
POST /folders {"title":"Q2","color":"#38bdf8"} → 200.
POST /tags {"title":"VIP","color":"#10b981"} → 200.
Frontend:
- createFolder / createTag clients accept an optional color param.
- LabelListModal now picks a default palette color when entering
add-row mode (rotating with item count) and offers a swatch
popover to override before submitting. Selected color is sent to
the backend.
CAMPAIGN_SELECT_FULL had:
array_agg(cet.tag_id) FILTER (WHERE cet.tag IS NOT NULL)
array_agg(cec.folder_id) FILTER (WHERE cec.folder IS NOT NULL)
The columns referenced in the FILTER clauses don't exist:
warmbly_dev=# \d campaign_email_tags
Column | Type
-----------+------
tag_id | uuid
campaign_id | uuid
warmbly_dev=# \d campaign_folders
Column | Type
-----------+------
campaign_id | uuid
folder_id | uuid
Result: every GET /campaigns returned 500 with
*pgconn.PgError: ERROR: column cet.tag does not exist (SQLSTATE 42703)
which is why the frontend page was perpetually blank — the request was
failing before any data could land. Fixed both FILTER predicates to
use the actual *_id columns.
Verified after rebuild:
- dev@warmbly.com (no campaigns): 200 with empty data array.
- beth@beta.test (owns seeded campaign): 200 with the Beta Cold
Outreach Q1 record.
1) Contacts crash "c is null":
contactRepository.Search declared `var contacts []models.Contact` so
an empty result set returned a nil slice, which Go marshals as JSON
null. The frontend's flatMap((p) => p.data) over null yields [null],
and the page then accesses c.subscribed → throws. Initialize as
make([]models.Contact, 0, limit+1) so the wire format is always [].
Also defensive on the client: useSearchContacts + useCampaigns now
coerce p.data ?? [] and drop nulls before returning.
2) Campaigns panic on any non-empty result:
campaignRepository.Search allocated `make([]models.Campaign, 0, limit+1)`
(length 0) then did `campaigns[i] = campaign`. That's an
index-out-of-range on the first iteration. Switched to `append`.
Anyone with at least one campaign would see a 500 / blank screen.
3) Websocket "Token expired":
SocketTTL was 60s. The frontend reconnect backoff caps at 30s, so
after a rejected handshake the next attempt could fire 30-60s
later. Combined with rare back-pressure on /getaway the token was
already past exp by the time the realtime saw it. Bumped to 10 min
— short enough to keep the token low-impact, long enough to outlast
the backoff schedule.
4) Websocket "Connection limit exceeded":
Realtime.Connections only untracked on channel terminate, never on
socket disconnect. Sockets that connected and disconnected without
joining a channel leaked. Each reconnect loop bumped the counter
until the per-user limit (10) was hit, after which every legitimate
connect was rejected even after fixing #3.
Fix: GenServer Process.monitor's the socket pid on track, and
`:DOWN` handler calls do_untrack with the right (user_id, ip).
5) Phoenix protocol mismatch:
Frontend appended vsn=2.0.0 to the WS URL, but sendRaw + joinChannel
send the V1 object format. Realtime's Phoenix.Socket.V2.JSONSerializer
crashed with a badmatch on the first phx_join, killing the socket
right after connect. Switched to vsn=1.0.0 to match what the client
actually emits.
Backend root cause:
The frontend client omits ?limit= when it would equal the default
(DEFAULT_PAGINATION_LIMIT = 50). validate.Limit("") treated empty as
invalid and returned errx.ErrLimit → 400 on /contacts/search and
/campaigns. ContactsTable derived isLoading from `!contacts`, which
stays true forever when the query errors, so the page hung in the
skeleton state instead of surfacing the error.
Fix:
- validate.Limit now accepts "" and returns LimitDefault = 50, in
sync with the frontend constant. The frontend's omission semantics
("don't send the param when it's the default") was already correct;
it was the validator that was wrong.
Frontend:
- ContactsTable: use isPending/isError/refetch directly from react-query
instead of deriving from `contacts`. New explicit error block renders
inside the body with: red alert tile, server error message, Try-again
button (with spinner during refetch), and Reload-page fallback.
- Campaigns page: same error UI promoted from the old EmptyBlock CTA
to a prominent block — alert tile + message + retry + reload.
New ContactFilters sheet (was the legacy 800px poppins drawer):
- 420px right-side panel matching the rest of the theme.
- Sticky 48px header with "Filters · N active" eyebrow + close.
- Sticky 48px footer with Reset / Cancel / Apply (slate-900 primary).
- Hairline-divided SectionBars between groups: Search, Custom field
filters, Sort, Subscription, Campaign membership, Dates.
- Custom field rows pair TextInput + FILTER_TYPES popover + value
input + remove button — all 28px tall.
- Sort: SelectButton popover + asc/desc toggle.
- Subscription: 3-state pill toggle (Any / Subscribed / Unsubscribed).
- Min/max campaign rows: checkbox toggle + number input + suffix.
- Date rows: checkbox toggle + native date input.
- Draft state mirrors parent until Apply, so editing filters doesn't
trigger refetches mid-build.
Root cause for the 10-min auto-logout (confirmed via pg_stat_activity):
the postgres pool MaxConns was 4, and four repository functions opened
a tx without committing or rolling back. After four calls each leaked
a connection in "idle in transaction" state. Once all four were gone
the pool was permanently exhausted — every new request that needed a
connection blocked until the client gave up. The 10-min trigger is
because that's when the first /auth/refresh fires; refresh tries to
acquire a connection, hangs, eventually the browser aborts the request,
the frontend treats the failure as session expiry, kicks the user.
The four leaking sites:
- emailRepository.Search (drove the leak — Accounts page)
- campaignRepository.Search
- sequenceRepository.Create
- contactRepository.BulkUpdate
Each now has `defer tx.Rollback(ctx)` immediately after Begin, matching
the pattern used in the non-leaky sites in the same files. Rollback is
a no-op after Commit, so this is safe for both read-only tx (Search)
and read-write tx (Create / BulkUpdate).
Additional hardening so a future leak can't silently brick the backend:
- MaxConns 4 → 25. 4 was reckless even without leaks; one bursty
admin page would saturate. 25 is still well under postgres'
default max_connections=100.
- MinConns 0 → 2. Keep a couple of warm connections at idle so the
first request after a quiet period doesn't pay the connect cost.
- idle_in_transaction_session_timeout=300000 (5 min) as a session
RuntimeParam. If a code path forgets the defer, postgres aborts
the leaked tx after 5 min and reclaims the connection.
- statement_timeout=60000 (60 s) as a session RuntimeParam.
Statement runaway can't pin a connection forever.
Verified after backend restart:
SELECT count(*) FROM pg_stat_activity
WHERE datname='warmbly_dev' AND state='idle in transaction';
→ 0
After /auth/refresh, Postgres got the new access + refresh nonces but
the Redis cached session still held the OLD ones. The next request:
1. Frontend uses the new access token (new access_nonce in JWT)
2. Backend ValidateAccessToken → GetSession → hits Redis cache
3. Cached session has the OLD access_nonce
4. session.AccessNonce != t.Nonce → ErrToken (401)
5. Frontend tries to refresh with the new refresh token
6. RefreshToken → GetSession → again hits stale Redis
7. sess.RefreshNonce (old) != t.Nonce (new) → ErrToken
8. Frontend clears tokens and bounces to /auth/login
The access token's 10-minute TTL was the trigger window because that's
when the first refresh fires. After the first refresh, the stale cache
poisoned every subsequent request.
Fix: delete the cached session after a successful repository update,
mirroring what SwitchOrganization already does for the same reason
(it updates current_organization_id in Postgres and then drops the
Redis copy). Next GetSession misses, re-reads from Postgres, caches
the fresh nonces.
The deleteSession failure path is intentionally swallowed — the
refresh already succeeded and we returned the new tokens, so worst
case is the next request triggers another refresh, not a logout.
Migration 000032 + repo + endpoints for arbitrary string tags on
workers. The fixed attributes (worker_type, free_tier, risk_pool)
cover the dimensions assignment logic uses. Tags cover everything
else admins want to group by: region (eu-west, fra), provider
(hetzner, ovh), role (warmup-only, burst-capacity), customer cohort —
whatever.
Schema:
- worker_tags(worker_id, tag) composite PK
- tag VARCHAR(64), lowercase + dashed via CHECK constraint
- ON DELETE CASCADE so deleting a worker drops its tags
Endpoints:
- GET /admin/workers/tags list distinct tags (autocomplete)
- PUT /admin/workers/:id/tags replace tag set; normalizes input
Repo:
- GetWorkerTags / SetWorkerTags / ListAllWorkerTags
- HydrateWorkerTags batch-loads tags onto a slice of workers in one
round-trip so the dashboard list doesn't do N+1 queries
PUT is transactional (delete + bulk insert) so the list view never
catches a worker mid-tag-swap. Auto-derived "smart" labels
(tier:free, pool:risky, state:error) are NOT stored — those are
computed client-side from the worker row so they stay in sync with
the source attributes automatically. Next commit wires the UI.
POST /admin/workers/preflight {host, port} runs a 5s TCP dial against
host:port and returns ok + latency, or an error. Used by the worker
creation wizard to catch typos / firewall problems while the form is
still open — much better UX than discovering an unreachable VPS at the
SSH test step after the row already exists.
Doesn't attempt an SSH handshake (no credentials at this stage). A green
preflight just means "something is listening there." The actual SSH test
runs later, after the admin pastes the generated pubkey.
Worker list grows a Pool column (clean=green, risky=amber,
quarantine=red badge). Dedicated workers render "n/a" — risk pools are
a shared-worker concept since dedicated workers don't share IPs across
customers.
Worker detail page (shared workers only) gets a "Risk pool" section
with three big buttons. Clicking a non-current pool confirms, then
calls PUT /admin/workers/:id/risk-pool. Action audited with the new
pool value.
Saving doesn't migrate accounts directly — the hourly rebalancer
notices the mismatch and moves mailboxes to a matching-pool worker
on its next tick. Documented in the section's helper text.
Endpoint accepts {risk_pool: "clean"|"risky"|"quarantine"} and is
gated by AdminPermManageWorkers. The worker detail row scan now
includes risk_pool so the column actually has data.
New background job in the consumer process:
1. Pulls up to 1000 mailbox candidates joined with their worst warmup
health state (across all pools they participate in) and their
current worker's risk_pool. Dedicated workers are excluded — single
tenant, segregation not applicable.
2. Recomputes risk_band from health state via RiskBandFromHealth.
If it changed, writes the new band.
3. If the band's matching pool doesn't equal the worker's pool, picks
a new worker via SelectSharedWorkerForBand and migrates the mailbox.
Increments/decrements account counts.
4. Logs each migration to admin_audit_log with action=
"risk_rebalance_migrate" so operators see what moved and why.
Boot-time run + hourly ticker. Rebalancing is intentionally batch, not
event-driven: warmup health states change on a slow rolling-window basis
(warmup_health_sweep is also hourly), so reacting in real time gains
nothing and would cause thundering-herd migrations.
JobsService gets an AssignmentService dep. Nil disables the job (lets
self-hosters opt out by simply not wiring it).
New method on WorkerAssignmentService picks the least-loaded shared
worker whose risk_pool matches the mailbox's risk band. Three-step
fallback chain:
1. Exact match: same pool, same tier
2. Fall back to clean pool of the same tier when no matching-pool
worker is available (better to land risky mailboxes on clean
workers than refuse; the rebalancer will move them later)
3. Last resort: any worker of the right tier (preserves legacy
behavior for installations that haven't provisioned risky/
quarantine pools)
Existing SelectSharedWorker is unchanged so call sites that don't
know about risk bands keep working. The next commit (background
rebalancer) is the first consumer of the new method.
Threat-level segregation, schema layer. Two new concepts:
workers.risk_pool ∈ {clean, risky, quarantine}
buckets shared workers by acceptable risk. Dedicated workers don't
use it (single tenant = no cross-contamination risk).
email_accounts.risk_band ∈ {clean, risky, quarantine}
per-mailbox classification, derived from warmup_health_state by the
rebalancer (next commit). Never set by user input.
The mapping is one-way and intentionally simple:
healthy → clean
watch, throttled → risky
quarantined, → quarantine
blocked
Rebalancer code lands in the next commit. This commit just adds:
- migration 000031 with enums + columns + filtered indexes
- WorkerRiskPool / EmailRiskBand types + RiskBandFromHealth helper
- WorkerRepository methods: SetWorkerRiskPool, SetEmailAccountRiskBand,
GetSharedWorkersByTierAndPool, ListRiskCandidates
- RiskCandidate result type joining email_accounts + warmup health
(picks WORST state across pools via CASE ranking) + worker columns
so the rebalancer can decide migrations in one scan
POST /admin/workers/:id/convert-to-dedicated does three things in
sequence:
1. Drain existing accounts to a supplied drain_to_worker_id (required
if the source has any accounts; we don't auto-pick per-account
targets because the right choice depends on each account's
owning org).
2. Flip workers.worker_type from "shared" to "dedicated".
3. Atomic create of dedicated_worker_assignments binding the worker
to a specific user/subscription (uses the existing
CreateDedicatedAssignmentIfNotExists so re-running is safe).
Refusal cases:
- already dedicated → 400
- has accounts but no drain target → 400 (admin must pick where they go)
- drain target equals source → 400
Worker detail UI gains a "Convert to dedicated" section, shown only
when the worker is currently shared. Inline form, no modal. The drain
dropdown excludes self, only lists shared+installed workers, sorts
least-loaded first.
Audit-logged with action="convert_to_dedicated" and the user_id,
subscription_id, drain target, and account count in details.
The free-tier-vs-paid separation in AssignWorkerToEmail is one of those
rules that's silently load-bearing: if a free org ever slips onto a
premium worker, the IPs of paying customers absorb the deliverability
hit. The code is correct today (strict isPaidOrg check at line 67, free/
premium pool sync at line 116), but nothing was guarding against a
regression.
Five table-thin tests, hand-rolled stub repos (embed the interface as a
nil field so unused methods panic loudly):
- free org → free shared worker → free warmup pool
- paid org → premium shared worker → premium warmup pool
- paid org with DedicatedWorkers > 0 + an assignment → dedicated worker
- paid org with DedicatedWorkers > 0 but no assignment → falls back to
premium shared (not free!)
- SelectSharedWorker with no workers → ErrNoAvailableWorkers
No code changes — this commit is documentation.
When the dead-worker job reassigns email accounts from a worker whose
heartbeat expired, write a row into admin_audit_log so the dashboard's
audit viewer surfaces these system actions alongside admin-driven ones.
admin_user_id is uuid.Nil (the platform identity), so admins searching
the log can distinguish "system did this" from "an admin did this" by
filtering on that ID. Details include the replacement worker, account
count, and reason.
JobsService gets an optional AdminRepo dep. Nil disables logging — keeps
the contract loose for any other call site that doesn't have one.
The audit calls added in the last commit went to AuditService.LogAction,
which writes to the general user-facing audit log (Cassandra). The admin
audit-log viewer at /admin/audit-logs queries the admin_audit_log table
in Postgres, so worker / credentials / release actions never showed up.
Add a public AdminService.LogAdminAction that wraps the existing private
logAction (writes to admin_audit_log with the same shape as ban_user /
update_worker / etc.). Repoint h.audit() at it.
Actions now visible in the audit viewer:
test, install, restart, upgrade, uninstall, rotate_keys, apply,
assign, system_update, reboot, check_releases (plus the existing
create/update/delete across workers, AWS creds, and profiles).
Matthew Meszaros