diff --git a/services/mcp-server/src/lib/extraction-client.ts b/services/mcp-server/src/lib/extraction-client.ts
index e4809be8..ae4351bb 100644
--- a/services/mcp-server/src/lib/extraction-client.ts
+++ b/services/mcp-server/src/lib/extraction-client.ts
@@ -192,12 +192,12 @@ export async function extractionGetProductRateLimitStatus(
 
 export async function extractionResetProductRateLimit(
   productId: string,
-  opts: { requestId?: string }
+  opts?: { requestId?: string }
 ): Promise<unknown> {
   const url = `${config.EXTRACTION_SERVICE_URL}/api/extract/rate-limits/product/reset`;
   const headers: Record<string, string> = {
     'Content-Type': 'application/json',
-    ...(opts.requestId ? { 'x-request-id': opts.requestId } : {}),
+    ...(opts?.requestId ? { 'x-request-id': opts.requestId } : {}),
   };
   const res = await fetch(url, {
     method: 'POST',
diff --git a/services/mcp-server/src/modules/a2a/post-incident-cleanup-pipeline.ts b/services/mcp-server/src/modules/a2a/post-incident-cleanup-pipeline.ts
index 79ec304b..2624652f 100644
--- a/services/mcp-server/src/modules/a2a/post-incident-cleanup-pipeline.ts
+++ b/services/mcp-server/src/modules/a2a/post-incident-cleanup-pipeline.ts
@@ -148,17 +148,30 @@ export async function runPostIncidentCleanupPipeline(
           continue;
         }
 
-        await telemetryDeletePolicy(policy.id, {
-          token: req.headers.authorization?.replace('Bearer ', '') || '',
-          requestId: req.id,
-        });
+        try {
+          await telemetryDeletePolicy(policy.id, {
+            token: req.headers.authorization?.replace('Bearer ', '') || '',
+            requestId: req.id,
+          });
 
-        policiesDeleted++;
+          policiesDeleted++;
 
-        req.log.info(
-          { runId, policyId: policy.id, policyName: policy.name },
-          'Deleted telemetry policy'
-        );
+          req.log.info(
+            { runId, policyId: policy.id, policyName: policy.name },
+            'Deleted telemetry policy'
+          );
+        } catch (policyError) {
+          req.log.warn(
+            {
+              runId,
+              policyId: policy.id,
+              policyName: policy.name,
+              error: policyError instanceof Error ? policyError.message : String(policyError),
+            },
+            'Failed to delete telemetry policy (continuing)'
+          );
+          // Continue with next policy instead of failing entire pipeline
+        }
       }
     }
 
@@ -169,25 +182,37 @@ export async function runPostIncidentCleanupPipeline(
         auditLogExported = true;
         auditLogSize = 0;
       } else {
-        const auditResponse = await telemetryQuery(
-          {
-            productId,
-            eventType: 'audit',
-            from: incidentTimeWindow.from,
-            to: incidentTimeWindow.to,
-            limit: 10000,
-          },
-          { token: req.headers.authorization?.replace('Bearer ', '') || '', requestId: req.id }
-        );
+        try {
+          const auditResponse = await telemetryQuery(
+            {
+              productId,
+              eventType: 'audit',
+              from: incidentTimeWindow.from,
+              to: incidentTimeWindow.to,
+              limit: 10000,
+            },
+            { token: req.headers.authorization?.replace('Bearer ', '') || '', requestId: req.id }
+          );
 
-        const auditEvents = auditResponse.events || [];
-        auditLogExported = true;
-        auditLogSize = JSON.stringify(auditEvents).length;
+          const auditEvents = (auditResponse as { events?: unknown[] }).events || [];
+          auditLogExported = true;
+          auditLogSize = JSON.stringify(auditEvents).length;
 
-        req.log.info(
-          { runId, eventCount: auditEvents.length, sizeBytes: auditLogSize },
-          'Exported audit log'
-        );
+          req.log.info(
+            { runId, eventCount: auditEvents.length, sizeBytes: auditLogSize },
+            'Exported audit log'
+          );
+        } catch (auditError) {
+          req.log.warn(
+            {
+              runId,
+              timeWindow: incidentTimeWindow,
+              error: auditError instanceof Error ? auditError.message : String(auditError),
+            },
+            'Failed to export audit log (continuing)'
+          );
+          // Continue without audit log instead of failing entire pipeline
+        }
       }
     }
 
diff --git a/services/mcp-server/src/modules/a2a/regression-watch-pipeline.ts b/services/mcp-server/src/modules/a2a/regression-watch-pipeline.ts
index 2ef28d91..0137a821 100644
--- a/services/mcp-server/src/modules/a2a/regression-watch-pipeline.ts
+++ b/services/mcp-server/src/modules/a2a/regression-watch-pipeline.ts
@@ -97,6 +97,7 @@ export async function runRegressionWatchPipeline(
     // Step 2: Create diagnostics sessions for representative clusters
     const sessionsToCreate = Math.min(clustersAboveThreshold.length, maxSessionsToCreate);
     const sessionIds: string[] = [];
+    let sessionsCreated = 0;
 
     for (let i = 0; i < sessionsToCreate; i++) {
       const cluster = clustersAboveThreshold[i];
@@ -107,40 +108,55 @@ export async function runRegressionWatchPipeline(
           '[DRY RUN] Would create diagnostics session'
         );
         sessionIds.push(`dry-run-session-${i + 1}`);
+        sessionsCreated++;
         continue;
       }
 
-      // Extract target from cluster - use fingerprint as identifier
-      const sessionResponse = await diagnosticsCreateSession(
-        {
-          productId,
-          targetUserId: 'system', // Use system user for cluster-based diagnostics
-          collectionLevel: 'trace',
-          captureLogs: true,
-          captureNetwork: true,
-          maxDurationMinutes: 30,
-        },
-        { token: req.headers.authorization?.replace('Bearer ', '') || '', requestId: req.id }
-      );
+      try {
+        // Extract target from cluster - use fingerprint as identifier
+        const sessionResponse = await diagnosticsCreateSession(
+          {
+            productId,
+            targetUserId: 'system', // Use system user for cluster-based diagnostics
+            collectionLevel: 'trace',
+            captureLogs: true,
+            captureNetwork: true,
+            maxDurationMinutes: 30,
+          },
+          { token: req.headers.authorization?.replace('Bearer ', '') || '', requestId: req.id }
+        );
 
-      sessionIds.push(sessionResponse.id);
+        sessionIds.push(sessionResponse.id);
+        sessionsCreated++;
 
-      req.log.info(
-        { runId, clusterId: cluster.id, sessionId: sessionResponse.id },
-        'Created diagnostics session for cluster'
-      );
+        req.log.info(
+          { runId, clusterId: cluster.id, sessionId: sessionResponse.id },
+          'Created diagnostics session for cluster'
+        );
+      } catch (sessionError) {
+        req.log.warn(
+          {
+            runId,
+            clusterId: cluster.id,
+            pk: cluster.pk,
+            error: sessionError instanceof Error ? sessionError.message : String(sessionError),
+          },
+          'Failed to create diagnostics session for cluster (continuing)'
+        );
+        // Continue with next cluster instead of failing entire pipeline
+      }
     }
 
     const summary = dryRun
       ? `[DRY RUN] Would create ${sessionsToCreate} diagnostics sessions for ${clustersAboveThreshold.length} clusters at ${severityThreshold} severity.`
-      : `Created ${sessionsToCreate} diagnostics sessions for ${clustersAboveThreshold.length} clusters at ${severityThreshold} severity.`;
+      : `Created ${sessionsCreated} diagnostics sessions for ${clustersAboveThreshold.length} clusters at ${severityThreshold} severity.`;
 
     return {
       runId,
       productId,
       clustersFound: clusters.length,
       clustersAboveThreshold: clustersAboveThreshold.length,
-      sessionsCreated: dryRun ? 0 : sessionsToCreate,
+      sessionsCreated: dryRun ? 0 : sessionsCreated,
       sessionIds,
       dryRun,
       summary,