apps/web/lib/tolt/schemas.ts (2)

44-44: Trim before slugifying to avoid leading/trailing dash artifacts.

Apply:

-  value: z.string().transform((val) => slugify(val)), // need to slugify this cause Tolt can sometimes return "john doe"
+  value: z.string().transform((val) => slugify(val.trim())), // need to slugify this cause Tolt can sometimes return "john doe"

---

50-54: Normalize empty strings to null for customer_id.
Empty string is not nullish; downstream ?? would pick it and produce an empty externalId.

Consider:

-export const ToltCustomerSchema = z.object({
+export const ToltCustomerSchema = z.object({
   id: z.string(),
-  customer_id: z
-    .string()
-    .describe("Internal customer identifier in client's app.")
-    .nullable(),
+  customer_id: z.preprocess(
+    (v) => (typeof v === "string" && v.trim() === "" ? null : v),
+    z.string().describe("Internal customer identifier in client's app.").nullable(),
+  ),
   email: z.string().nullable(),
   name: z.string().nullable(),

apps/web/lib/tolt/import-customers.ts (1)

160-161: Nit: include quotes around id for clearer logs.

-      `A customer already exists with customerExternalId ${customerExternalId}`,
+      `A customer already exists with customerExternalId "${customerExternalId}"`,

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

💡 Knowledge Base configuration:

• MCP integration is disabled by default for public repositories
• Jira integration is disabled by default for public repositories
• Linear integration is disabled by default for public repositories

You can enable these sources in your CodeRabbit configuration.

apps/web/lib/tolt/import-customers.ts (2)

127-133: Avoid empty-string entity_id; coalesce to first non-empty normalized value

Using ?? will accept "", causing bad keys and uniqueness collisions. Normalize email to lowercase and pick the first non-empty candidate.

-  const commonImportLogInputs = {
+  const normalisedEmail = customer.email?.trim().toLowerCase() ?? null;
+  const entityId =
+    [customer.customer_id, normalisedEmail, customer.id].find(
+      (v): v is string => !!v && v.trim().length > 0,
+    ) ?? customer.id;
+  const commonImportLogInputs = {
     workspace_id: workspace.id,
     import_id: importId,
     source: "tolt",
     entity: "customer",
-    entity_id: customer.customer_id ?? customer.email ?? customer.id,
+    entity_id: entityId,
   } as const;

---

145-148: Unify customerExternalId with normalized entityId

Reuse the computed entityId to keep logs, lookups, and persistence consistent and non-empty.

-  // if customer_id is null, use customer email or Tolt customer ID as the external ID
-  const customerExternalId =
-    customer.customer_id ?? customer.email ?? customer.id;
+  // external ID: first non-empty of customer_id, normalized email, or Tolt ID
+  const customerExternalId = entityId;

apps/web/lib/tolt/schemas.ts (1)

44-46: Trim before slugifying to avoid stray dashes and inconsistent keys

Slugify after a trim to normalize inputs like " john doe ". Also consider being explicit about lowercase for determinism.

-  value: z.string().transform((val) => slugify(val)), // need to slugify this cause Tolt can sometimes return "john doe"
+  value: z.string().transform((val) => slugify(val.trim())), // trim before slugify

apps/web/lib/tolt/import-customers.ts (1)

160-160: Use structured, low-PII logging

Prefer a concise, structured message and avoid dumping full objects where possible.

-    console.log(
-      `A customer already exists with customerExternalId ${customerExternalId}`,
-    );
+    console.info(`Customer exists; skipping import. externalId=${customerExternalId}`);

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

💡 Knowledge Base configuration:

• MCP integration is disabled by default for public repositories
• Jira integration is disabled by default for public repositories
• Linear integration is disabled by default for public repositories

You can enable these sources in your CodeRabbit configuration.