Yashas
cb94e2417f
The page said "Waiting for the customer to reply on WhatsApp" directly above a panel saying "Accepted by the customer, over WhatsApp. KYC and underwriting are running." Both rendered from the same instant; only one was true. Quote Presented has the same shape as Document Pending — the state holds still while the situation underneath it changes completely. Before the customer answers, the lead is genuinely with them and nothing is running. After they accept, KYC and the underwriting screen run inside the SAME state, because the lead does not move until underwriting clears. phaseOf splits it, so the strip now says "Verifying KYC" and then "Screening the risk for underwriting", and the Next step line records only the fact of the acceptance rather than repeating what is running. That also puts this stretch under the stall detector for the first time, which matters: the KYC agent talked itself out of the underwriting screen earlier today and the lead sat in Quote Presented with a passed KYC and nothing happening. With this it would have said so after five minutes and offered "Run the underwriting screen" — both steps ops_admin can perform, so the recovery needs nobody's help. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|---|---|---|
| public/brand | ||
| src | ||
| .env.example | ||
| .gitignore | ||
| CLAUDE.md | ||
| eslint.config.js | ||
| index.html | ||
| package-lock.json | ||
| package.json | ||
| README.md | ||
| vite.config.js | ||
Zurich Kotak — Lead Desk
Operator console for the Zurich Kotak "Lead to Policy" demo. Leads arrive through three channels — direct, bancassurance and agency — and run one state machine to policy issuance. The agents do the work; a person appears at one queue.
Workflow config is seeded from sm2/custom-apps/zurich-kotak/ (org 84,
app 536, workflow zk_wf_lead). Design doc:
sm2/app-designs/zurich-kotak/design.html. Workflow spec PDF:
sm2/custom-apps/zurich-kotak/zurich-kotak-workflow.pdf.
Run it
npm install
npm run dev # http://localhost:5175
npm run build # → dist/
.env carries the API host for local dev only:
VITE_ZINO_API_URL=https://dev.getzino.in
Logins (all password ZurichKotak@2026, org 84):
| Role | Can do | |
|---|---|---|
sanjay.uw@zurichkotak.example |
SME Underwriter | Clear / Decline Referral — the only human decision |
meera.rm@zurichkotak.example |
Bancassurance RM | Submit a Partner Bank Lead |
arjun.posp@zurichkotak.example |
POSP / Broker | Submit a Partner Agent Lead |
kavya.csr@zurichkotak.example |
Direct / Call centre | Submit a Self-Serve Lead |
deepa.ops@zurichkotak.example |
Operations | Cross-channel visibility |
Sign in as Sanjay to land in Referred to Underwriting, which is where the demo happens.
How it talks to the platform
Everything goes through src/api/client.js:
POST /usr/login— auth.org_idmust be a string; a number returnscannot unmarshal number into Go struct field LoginRequest.org_id.POST /app/536/view/recordview— every queue is one record view (zk-rv-leads) filtered server-side oncurrent_state_name.POST /app/536/view/form-screensthenPOST /app/536/activity— forms are read from the live activity schema and submitted straight back.
Forms are not defined in this repo
Whatever /view/form-screens returns is what renders — labels, types, select
options, which fields are mandatory. Add a field to an activity in Studio,
redeploy, and it appears here with no frontend change. That is deliberate: the
workflow is the source of truth, and a hardcoded form would quietly diverge
from it.
Permissions are never enforced here
The console does not hide a button to stop someone using it. Every permission
decision is the workflow's, made server-side on each submission — the platform
refuses and this app reports what it said. Two refusals worth knowing:
/activity answers 403, /start answers 400, and both carry
permission denied.
The runtime-config contract
VITE_ZINO_API_URL is read at runtime from a config.js the server writes
when it places the build — never compiled in. One artifact is promoted between
environments unchanged, so a build-time URL would point every environment at
whichever backend happened to build it. requireConfigValue throws if it is
missing, so a production build with no config.js fails loudly rather than
calling the wrong backend.
vite.config.js uses base: './' and the router takes its basename from the
<base href> the server writes, so one build serves any mount path. Do not
reintroduce a build-time base.
This is also why the fonts and logo live under src/assets/ and not
public/. Vite rewrites bundled asset URLs to be relative; a public/ file
referenced as /fonts/… stays absolute and 404s under the /zurich-kotak/
mount path. The favicon is the one exception — it stays in public/brand/ and
is referenced relatively from index.html.
Deployment
Registered with frontgen by sm2/custom-apps/zurich-kotak/05_frontgen_project.sql
(slug zurich-kotak, repo_name zurich_kotak, org 84). A push to main builds
and serves at https://preview-dev.getzino.in/zurich-kotak/.
The webhook only fires on a new push — a push made before the frontgen project row existed matched no project and did nothing.
State of the build
| Piece | Status |
|---|---|
| Brand — Zurich Sans, palette, logo | done (kept from the original scaffold) |
| Sign-in against the real gateway | done |
| Runtime config, relative base, router | done |
Pipeline sidebar (mirrors tbl_wf_states) |
done |
| Queue lists | waiting on the zk-rv-leads record view |
| Lead file, activity forms, attribution panel | waiting on zk-dv-lead + form screens |