Admin access tokens expire after 15 minutes with no refresh flow. Until now,
once the token expired, every subsequent action (approve, save, etc.) just
403/401'd with no visible feedback, or an error banner easy to miss inside a
modal — it looked like the button "didn't work."
Add a global window.fetch wrapper (installSessionExpiryWatcher, installed
once from AdminShell) that flips a shared signal whenever an authenticated
request comes back 401. AdminShell renders a "You have been logged out"
dialog on top of everything when that fires, with a button that clears the
stale session and sends the admin back to /login?from=<current path>.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Every action branch (reject, request-documents, request-revision, notes)
set common.body to a JSON payload, but the isApprove branch never did,
leaving body: undefined. axum's Json<ActionPayload> extractor rejects a
body-less POST before approve_verification even runs, so clicking Approve
in the queue-list modal silently failed (the error banner renders at the
top of the page, easy to miss behind the open modal).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Backend's AuthUser extractor only reads the Authorization header — it never
looks at cookies. The verification queue list (index.tsx load()) and every
call in the verification detail page (via the shared adminFetch helper) sent
credentials: 'include' but no Authorization header, so they 401'd right after
a successful login even though the access token was already in sessionStorage.
Matches the pattern already used by the approve/reject actions further down
in index.tsx.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Server-side gateway proxy (login, forgot-password, reset-password, all
/api/admin/* and /api/gateway/* calls) was reading import.meta.env.VITE_GATEWAY_URL,
which Vite only inlines at build time. The Dockerfile wrote GATEWAY_URL (no VITE_
prefix, and localhost besides) to .env at build time, so the var was never picked
up and the fallback http://localhost:9100 got baked into the server bundle
permanently — unreachable inside the pod, causing every admin auth call to 502
regardless of the correct GATEWAY_URL already set in the k8s ConfigMap.
Switch to process.env.GATEWAY_URL, read at runtime like the rest of the server
config already is.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Same fix as nxtgauge-frontend-solid: registry_prune.py sorted candidate
tags by the manifest GET response's Date header (the moment of the
request, not the image's actual build time), making the "keep newest N"
sort effectively random whenever multiple tags are touched in the same
prune run - which happens on every single run, since the tag just
pushed by this same build is always one of the candidates. That let the
prune step delete the image a run had just built, before gitops even
got a chance to reference it.
Now reads the real "created" timestamp from the image's config blob and
always protects the current run's own SHA from deletion regardless of
sort order.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The previous fix routed through /api/gateway/auth/* (the users-table
forgot-password), which succeeds but changes the wrong account entirely —
admin.nxtgauge.com authenticates against a separate employees table.
Backend now exposes /api/admin/auth/forgot-password and
/api/admin/auth/reset-password against EmployeeRepository; route there.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
/api/auth/forgot-password and /api/auth/reset-password aren't routes this
app serves itself — admin.nxtgauge.com only proxies specific prefixes
(/api/admin/*, /api/gateway/*, /api/me/*), so calling the bare path 404'd
at the SolidStart routing layer before ever reaching the backend. The
existing /api/gateway/[...path] proxy already rewrites /auth/* to
/api/auth/* on the real backend — route through that instead.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The reset-password flow called /api/auth/internal/forgot-password/request-code
and /verify-code with a challengeId-based two-step contract — neither route
exists anywhere in the backend, so every reset attempt failed with "Failed to
send reset code." The real backend only exposes /api/auth/forgot-password
(sends a one-time code, always 200 to avoid account enumeration) and
/api/auth/reset-password (consumes the code + new password, no challengeId
involved). Rewrites both request/verify handlers to match that contract.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Each of these profession admin list pages (makeup artist, tutor, developer,
video editor, graphic designer, social media manager, fitness trainer,
catering services) linked "View Profile" to /admin/{role}/{id}, but only
Photographer and Company had a real [id].tsx route — every other role fell
through to the legacy catch-all module and showed a placeholder instead of
the applicant's actual profile. Converts each flat list route into a
folder (index.tsx + [id].tsx) and adds a shared ProfessionDetailPage
component, modeled on the working Photographer detail page.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
- users: wire suspend/block buttons to PATCH /api/admin/users/:id/status
(previously only mutated local state, reverted on refresh); add
load/action error banners instead of silently emptying the table
- [...module]: stop rendering a dead localhost:9201 iframe for legacy
modules when VITE_LEGACY_ADMIN_ORIGIN isn't configured (it never is
in production); show a clear "not available" message instead
- dashboard: surface a banner when /api/admin/dashboard/metrics fails
instead of silently showing "No Data" on every widget
- credit: fix stray extra closing </Show> tag that broke the whole
file's JSX parse; restore missing API/authHeaders module helpers
dropped in a previous refactor (AI Credits handlers referenced them
but they were undefined); replace a dead, broken exportLedgerCsv/
filteredLedger implementation with one matching actual call sites;
fix activeTab type/tab keys so the Balance & Ledger and Platform
Ledger tabs were actually reachable (they compared against 'balance'
/'platform' but the tab buttons only ever set 'ledger')
- roles: surface errors when fetching a role's permissions for edit
fails, instead of silently swallowing them
- runtime-roles: surface fetch/delete errors instead of silently
swallowing them or presenting fallback sample data as if real
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
GITOPS_PAT auth was failing against Forgejo. GitHub is the source of
truth for nxtgauge-gitops; push there and let sync-to-forgejo.yml
relay to Forgejo, which Flux actually watches.
Same fix as nxtgauge-frontend-solid: GITEOPS_REPO/GITEOPS_SSH_KEY were
never configured as repo secrets. Switch to an HTTPS clone with
GITOPS_PAT, matching nxtgauge-backend-rust's already-working workflow.
Same fix as nxtgauge-frontend-solid: buildx writes its metadata file
pretty-printed, so the compact ":"-only grep pattern never matched,
failing the build-and-push step right after a successful push and
skipping the GitOps-release step. Match backend-rust's
whitespace-tolerant pattern.
Same fix as nxtgauge-backend-rust and nxtgauge-frontend-solid: every
build here has been failing with "Cannot connect to the Docker daemon
at tcp://127.0.0.1:2375" since the job container is nested inside the
runner pod's dind sidecar. Detect the real gateway from
/proc/net/route instead of hardcoding loopback.
- verification/[id].tsx: "Add Note" in the Reviewer Notes panel had no
onClick — wired to the existing POST /api/admin/verifications/:id/notes
endpoint.
- approval/index.tsx: Decision Notes textarea was uncontrolled and its
"Add Note" button was a no-op; there's no standalone notes endpoint
for approval_requests, so the textarea is now bound to state and its
value is sent as the rejection reason when the reviewer rejects
(the only channel the backend currently supports).
- support.tsx: removed a no-op onClick={() => {}} on the case row —
the row isn't actually clickable, only the adjacent "View" link is
(though that link's target route doesn't exist yet — flagged
separately, out of scope here).
- Remove build-and-deploy-forgejo.yml from GitHub Actions
- Forgejo CI at ci.nxtgauge.com now handles all builds and deployments
- All images pushed to Forgejo registry (ci.nxtgauge.com/ashwin/*)
- Update workflow to push to ci.nxtgauge.com/ashwin/nxtgauge-admin-solid
- Use FORGEJO_USERNAME and FORGEJO_TOKEN secrets
- Fix gitops repo cloning to use x-access-token
- Fix nested tag structure: AI Credits tab was incorrectly nested inside Platform Ledger section
- Close Platform Ledger section before starting AI Credits tab
- Move AI Credits tab to be a sibling of Platform Ledger tab, not a child
- Wrap AI Credits content in proper section element for consistent styling
Fixes build errors in admin-solid
- Fix nested tag structure: AI Credits tab was incorrectly nested inside Platform Ledger section
- Close Platform Ledger section before starting AI Credits tab
- Move AI Credits tab to be a sibling of Platform Ledger tab, not a child
- Wrap AI Credits content in proper section element for consistent styling
- Inject Bearer token from sessionStorage into all API requests
- Add typed API helpers for invoices, credits, discounts, ledger, orders, pricing, tax
- Overhaul invoice, credit, pricing, tax admin routes with full CRUD UI
- Remove committed log files
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
New AI Management page (/admin/ai-management):
- Overview tab: AI users, daily/monthly generations, active plans
- User Usage tab: searchable table with plan, usage, addon balance
- Plans tab: grid of available AI plans
Sidebar:
- Added AI Management nav item with Sparkles icon
Dashboard:
- Added kpi_ai_users widget
- Added kpi_ai_generations_today widget
Files:
- src/routes/admin/ai-management.tsx (new)
