fix: address PR #759 round-2 review feedback

- jira-settings / linear-settings: drop the local STATUS_REFRESH_MS = 90_000
  copies and import the shared INTEGRATION_STATUS_REFRESH_MS so changing the
  cadence only touches one file (Claude).
- healthpoll: rename Poller.ProbeAll → probeAll. The only callers are the
  loop and the in-package tests, so widening the public API wasn't buying
  anything (Claude).

apps/backend/internal/integrations/healthpoll/healthpoll.go

@@ -103,7 +103,7 @@ func (p *Poller) loop(ctx context.Context) {
 	defer p.wg.Done()
 	// Run an initial probe immediately so the UI gets a status without
 	// waiting the full interval after backend startup.
-	p.ProbeAll(ctx)
+	p.probeAll(ctx)
 	ticker := time.NewTicker(p.interval)
 	defer ticker.Stop()
 	for {
@@ -111,14 +111,14 @@ func (p *Poller) loop(ctx context.Context) {
 		case <-ctx.Done():
 			return
 		case <-ticker.C:
-			p.ProbeAll(ctx)
+			p.probeAll(ctx)
 		}
 	}
 }
 
-// ProbeAll runs one probe pass across every configured workspace. Exposed so
-// tests can drive it deterministically without having to wait for a tick.
-func (p *Poller) ProbeAll(ctx context.Context) {
+// probeAll runs one probe pass across every configured workspace. Used by the
+// loop and by in-package tests that want to drive a probe deterministically.
+func (p *Poller) probeAll(ctx context.Context) {
 	ids, err := p.prober.ListConfiguredWorkspaces(ctx)
 	if err != nil {
 		p.logger.Warn(p.name+" poller: list workspaces failed", zap.Error(err))

apps/backend/internal/integrations/healthpoll/healthpoll_test.go

@@ -46,7 +46,7 @@ func TestProbeAll_RecordsEachWorkspace(t *testing.T) {
 	}
 	poller := New("test", p, logger.Default())
 
-	poller.ProbeAll(context.Background())
+	poller.probeAll(context.Background())
 
 	p.mu.Lock()
 	defer p.mu.Unlock()
@@ -63,7 +63,7 @@ func TestProbeAll_ListError_DoesNotPanic(t *testing.T) {
 	}
 	poller := New("test", p, logger.Default())
 
-	poller.ProbeAll(context.Background()) // expect: silent recovery + warning log
+	poller.probeAll(context.Background()) // expect: silent recovery + warning log
 
 	p.mu.Lock()
 	defer p.mu.Unlock()
@@ -82,7 +82,7 @@ func TestProbeAll_StopsWhenContextCancelled(t *testing.T) {
 
 	ctx, cancel := context.WithCancel(context.Background())
 	cancel() // cancel before the loop runs
-	poller.ProbeAll(ctx)
+	poller.probeAll(ctx)
 
 	p.mu.Lock()
 	defer p.mu.Unlock()

apps/web/components/jira/jira-settings.tsx

@@ -19,6 +19,7 @@ import {
   IntegrationAuthStatusBanner,
   type IntegrationAuthHealth,
 } from "@/components/integrations/auth-status-banner";
+import { INTEGRATION_STATUS_REFRESH_MS } from "@/hooks/domains/integrations/use-integration-availability";
 import {
   getJiraConfig,
   setJiraConfig,
@@ -27,10 +28,6 @@ import {
 } from "@/lib/api/domains/jira-api";
 import type { JiraAuthMethod, JiraConfig, TestJiraConnectionResult } from "@/lib/types/jira";
 
-// How often to re-fetch the config so the auth-health indicator picks up new
-// probe results from the backend poller (which runs every 90s).
-const STATUS_REFRESH_MS = 90_000;
-
 // Session cookies are HttpOnly so document.cookie can't read them, but
 // DevTools → Application → Cookies surfaces them in plain text. Users copy
 // the Value cell of a single row; the backend wraps it under both
@@ -368,7 +365,7 @@ function useJiraSettings(workspaceId: string) {
         .catch(() => {
           /* transient failures are fine — next tick retries */
         });
-    }, STATUS_REFRESH_MS);
+    }, INTEGRATION_STATUS_REFRESH_MS);
     return () => clearInterval(id);
   }, [workspaceId]);
 

apps/web/components/linear/linear-settings.tsx

@@ -17,6 +17,7 @@ import {
   IntegrationAuthStatusBanner,
   type IntegrationAuthHealth,
 } from "@/components/integrations/auth-status-banner";
+import { INTEGRATION_STATUS_REFRESH_MS } from "@/hooks/domains/integrations/use-integration-availability";
 import {
   getLinearConfig,
   setLinearConfig,
@@ -26,10 +27,6 @@ import {
 } from "@/lib/api/domains/linear-api";
 import type { LinearConfig, LinearTeam, TestLinearConnectionResult } from "@/lib/types/linear";
 
-// How often to re-fetch the config so the auth-health indicator picks up new
-// probe results from the backend poller (which runs every 90s).
-const STATUS_REFRESH_MS = 90_000;
-
 type FormState = {
   defaultTeamKey: string;
   secret: string;
@@ -336,7 +333,7 @@ function useLinearSettings(workspaceId: string) {
         .catch(() => {
           /* transient failures are fine — next tick retries */
         });
-    }, STATUS_REFRESH_MS);
+    }, INTEGRATION_STATUS_REFRESH_MS);
     return () => clearInterval(id);
   }, [workspaceId]);