fix(build): retry transient sitemap endpoint fetches

The TypeScript helper refactor exposed a separate production risk: builds can fail when the sitemap-endpoints API returns a transient 5xx during Pagefind index generation. Move that fetch into a small helper with retry logic and add a focused prebuild test so this failure mode is caught without waiting on a full deploy.

Constraint: Must keep Pagefind index generation behavior the same when the endpoint is healthy
Rejected: Ignore the failure as external-only | transient 5xx responses can still block deploys and CI repeatedly
Confidence: high
Scope-risk: narrow
Reversibility: clean
Directive: When external build inputs can fail transiently, add a small retryable helper plus a direct test before relying on full deploy verification
Tested: pnpm exec vitest run test/prebuild/load-sitemap-endpoints.test.ts; pnpm run typecheck; pnpm run test-prebuild
Not-tested: Full production redeploy completion before push

helpers/pagefind/load-sitemap-endpoints.ts

@@ -0,0 +1,63 @@
+import fs from 'fs-extra'
+import axios from 'axios'
+
+import {
+    sitemapEndpointsPath
+} from '~/helpers/pagefind/config.js'
+
+function shouldRetryError ( error: unknown ) {
+    const status = ( error as { response?: { status?: number } } )?.response?.status
+
+    return typeof status === 'number' && status >= 500
+}
+
+async function fetchJsonWithRetries (
+    url: string,
+    {
+        attempts = 3,
+        delayMs = 1000
+    }: {
+        attempts?: number
+        delayMs?: number
+    } = {}
+) {
+    let lastError: unknown
+
+    for ( let attempt = 1; attempt <= attempts; attempt += 1 ) {
+        try {
+            const response = await axios.get( url )
+
+            return response.data
+        } catch ( error ) {
+            lastError = error
+
+            if ( attempt >= attempts || !shouldRetryError( error ) ) {
+                throw error
+            }
+
+            await new Promise( resolve => setTimeout( resolve, delayMs ) )
+        }
+    }
+
+    throw lastError
+}
+
+export async function loadSitemapEndpoints () {
+    if ( await fs.pathExists( sitemapEndpointsPath ) ) {
+        return await fs.readJson( sitemapEndpointsPath )
+    }
+
+    if ( !process.env.PUBLIC_API_DOMAIN ) {
+        throw new Error(`Missing ${ sitemapEndpointsPath } and PUBLIC_API_DOMAIN is not set`)
+    }
+
+    const apiUrl = new URL( process.env.PUBLIC_API_DOMAIN )
+    apiUrl.pathname = sitemapEndpointsPath.replace(/^\.?\/?static\//, '/')
+
+    return await fetchJsonWithRetries( apiUrl.toString() )
+}
+
+export {
+    fetchJsonWithRetries,
+    shouldRetryError
+}

scripts/build-pagefind-index.js

@@ -1,31 +1,12 @@
-import fs from 'fs-extra'
-import axios from 'axios'
 import 'dotenv/config.js'
 
 import {
-    sitemapEndpointsPath
-} from '~/helpers/pagefind/config.js'
+    loadSitemapEndpoints
+} from '~/helpers/pagefind/load-sitemap-endpoints'
 import {
     writePagefindIndex
 } from '~/helpers/pagefind/index.js'
 
-async function loadSitemapEndpoints () {
-    if ( await fs.pathExists( sitemapEndpointsPath ) ) {
-        return await fs.readJson( sitemapEndpointsPath )
-    }
-
-    if ( !process.env.PUBLIC_API_DOMAIN ) {
-        throw new Error(`Missing ${ sitemapEndpointsPath } and PUBLIC_API_DOMAIN is not set`)
-    }
-
-    const apiUrl = new URL( process.env.PUBLIC_API_DOMAIN )
-    apiUrl.pathname = sitemapEndpointsPath.replace(/^\.?\/?static\//, '/')
-
-    const response = await axios.get( apiUrl.toString() )
-
-    return response.data
-}
-
 ;(async () => {
     const sitemapEndpoints = await loadSitemapEndpoints()
     const {

test/prebuild/load-sitemap-endpoints.test.ts

@@ -0,0 +1,94 @@
+import {
+    beforeEach,
+    describe,
+    expect,
+    it,
+    vi
+} from 'vitest'
+
+import axios from 'axios'
+
+import {
+    fetchJsonWithRetries,
+    shouldRetryError
+} from '~/helpers/pagefind/load-sitemap-endpoints'
+
+vi.mock( 'axios', () => {
+    return {
+        default: {
+            get: vi.fn()
+        }
+    }
+} )
+
+describe( 'load sitemap endpoints helper', () => {
+    beforeEach( () => {
+        vi.mocked( axios.get ).mockReset()
+    } )
+
+    it( 'retries transient 5xx errors and eventually resolves', async () => {
+        const axiosGet = vi.mocked( axios.get )
+
+        axiosGet
+            .mockRejectedValueOnce({
+                response: {
+                    status: 502
+                }
+            })
+            .mockResolvedValueOnce({
+                data: {
+                    ok: true
+                }
+            } )
+
+        const data = await fetchJsonWithRetries( 'https://api.doesitarm.com/sitemap-endpoints.json', {
+            attempts: 2,
+            delayMs: 0
+        } )
+
+        expect( data ).toEqual({
+            ok: true
+        } )
+        expect( axiosGet ).toHaveBeenCalledTimes( 2 )
+    } )
+
+    it( 'does not retry non-5xx errors', async () => {
+        const axiosGet = vi.mocked( axios.get )
+
+        axiosGet.mockRejectedValueOnce({
+            response: {
+                status: 404
+            }
+        })
+
+        await expect( fetchJsonWithRetries( 'https://api.doesitarm.com/sitemap-endpoints.json', {
+            attempts: 3,
+            delayMs: 0
+        } ) ).rejects.toEqual({
+            response: {
+                status: 404
+            }
+        })
+
+        expect( axiosGet ).toHaveBeenCalledTimes( 1 )
+    } )
+
+    it( 'classifies retryable server errors', () => {
+        expect( shouldRetryError( {
+            response: {
+                status: 502
+            }
+        } ) ).toBe( true )
+        expect( shouldRetryError( {
+            response: {
+                status: 503
+            }
+        } ) ).toBe( true )
+        expect( shouldRetryError( {
+            response: {
+                status: 404
+            }
+        } ) ).toBe( false )
+        expect( shouldRetryError( new Error( 'network' ) ) ).toBe( false )
+    } )
+} )