Surface governed action posture in Patrol Assistant briefings

docs/release-control/v6/internal/subsystems/api-contracts.md

@@ -458,7 +458,10 @@ the canonical monitored-system blocked payload.
    finding entries may carry live pending Patrol approval posture only as safe
    structured handoff actions with approval ID/status/risk/target/request/expiry
    metadata, action plan identity/policy/expiry, dry-run posture, and command
-   counts instead of pasting page-local UI text or raw command payloads into chat
+   counts instead of pasting page-local UI text or raw command payloads into chat,
+   and may derive visible action labels, safety notes, and approval-aware
+   suggested prompts from that same safe metadata so the drawer does not fall
+   back to generic assessment copy when governed actions are present
    and that same Patrol investigation-context owner, so visible Assistant
    drawer handoffs may include live pending-approval metadata only as safe
    operator context: approval ID, status, risk, requested/expiry timestamps,
@@ -3037,10 +3040,12 @@ visible handoffs from findings, while the downstream chat
 service hydrates live resource state, timeline, and action audit context around
 that same handoff. Frontend Patrol handoff helpers may consume current pending
 approval list payloads only as safe metadata for that visible briefing and any
-structured `handoff_actions`: approval ID, status, risk, request/expiry
-timestamps, target label, action ID, approval policy, plan expiry, and dry-run
-summary are allowed, while approval command text remains inside the governed
-approval/remediation surface.
+   structured `handoff_actions`: approval ID, status, risk, request/expiry
+   timestamps, target label, action ID, approval policy, plan expiry, and dry-run
+   summary are allowed. Assessment-level visible briefings may reuse that same
+   safe metadata for action labels, safety notes, and approval-aware suggested
+   prompts, while approval command text remains inside the governed
+   approval/remediation surface.
 Patrol approval-row Assistant prompts must use the same safe metadata boundary
 and set `autonomousMode:false` for the request-local chat handoff; they must not
 paste raw approval or proposed-fix command text into the authored chat prompt.

docs/release-control/v6/internal/subsystems/frontend-primitives.md

@@ -819,7 +819,10 @@ frontend primitive boundary.
     approval policy, plan expiry, dry-run posture, and command count. Those
     entries may be passed through shared chat transport as `handoff_actions` for
     model-only refresh, but the shared drawer stays a generic shell rather than
-    a Patrol summary prompt builder. The
+    a Patrol summary prompt builder. The Patrol helper may turn those same safe
+    references into visible action labels, safety notes, and approval-aware
+    suggested prompts, but it must not expose raw command or execution payloads.
+    The
     drawer may render a generic
     context-briefing band from `frontend-modern/src/stores/aiChat.ts`, but
     feature-owned helpers must provide the source labels, attention reason,

docs/release-control/v6/internal/subsystems/patrol-intelligence.md

@@ -223,9 +223,10 @@ Patrol-specific presentation helpers.
    approval ID, pending status, risk, target, requested/expiry timestamps,
    action plan identity, approval policy, plan expiry, dry-run posture, and
    command count. They must force request-local approval-required mode, keep raw
-   command and approval payloads out of prompt and drawer copy, and frame
-   Assistant as explanation, prioritization, and safe next-step review rather
-   than a generic reactive chat
+   command and approval payloads out of prompt and drawer copy, surface visible
+   drawer action posture from the same safe references, and frame Assistant as
+   explanation, prioritization, and safe next-step review rather than a generic
+   reactive chat
    box.
 
 ## Current State
@@ -284,7 +285,10 @@ evidence, bounded active-finding summaries, source-owned suggested prompts, and
 deduped resource references plus safe structured approval/action references as
 model-only context while forcing `autonomousMode:false` and summarizing
 proposed-fix command-bearing records and command-bearing change events without
-raw command text.
+raw command text. Its visible Assistant briefing must also use those safe
+references to distinguish pending governed approvals or attached action
+references from a generic assessment discussion, including approval-policy and
+dry-run posture when available, while command payloads stay out of the drawer.
 That active-runtime label must stay operational rather than verdict-like: the
 header chip should communicate that Patrol is enabled or available, not imply
 that infrastructure health is currently good merely because the runtime is on.

frontend-modern/src/features/patrol/__tests__/PatrolIntelligenceSummary.test.tsx

@@ -88,10 +88,13 @@ describe('PatrolIntelligenceSummary', () => {
       actionDryRunSummary: 'No provider-supported dry run is available for this action.',
     });
     expect(context.briefing).toMatchObject({
+      actionLabel: '1 pending governed approval attached',
+      safetyNote:
+        'Review approvals in the governed flow; approval policy is attached; dry-run posture is attached; raw command payloads stay out of Assistant.',
       suggestedPrompts: [
         'Prioritize findings and safest next step',
         'Explain recent changes and correlations',
-        'Summarize governed remediation risks',
+        'Review pending approvals and safest next step',
       ],
     });
     expect(JSON.stringify(context)).not.toContain('systemctl restart workload.service');

frontend-modern/src/features/patrol/__tests__/patrolInvestigationContextModel.test.ts

@@ -262,7 +262,9 @@ describe('patrolInvestigationContextModel', () => {
       sourceLabel: 'Pulse Patrol',
       title: 'Patrol assessment attached',
       subject: 'Issues detected',
-      safetyNote: 'Diagnostics and remediation require governed approval.',
+      actionLabel: '1 governed action reference attached',
+      safetyNote:
+        'Review action posture in the governed flow; raw command payloads stay out of Assistant.',
       suggestedPrompts: [
         'Prioritize findings and safest next step',
         'Explain recent changes and correlations',
@@ -345,8 +347,13 @@ describe('patrolInvestigationContextModel', () => {
     expect(handoff.context.context).toMatchObject({
       pendingApprovalCount: 1,
     });
+    expect(handoff.context.briefing).toMatchObject({
+      actionLabel: '1 pending governed approval attached',
+      safetyNote:
+        'Review approvals in the governed flow; approval policy is attached; dry-run posture is attached; destructive actions remain approval-bound; raw command payloads stay out of Assistant.',
+    });
     expect(handoff.context.briefing?.suggestedPrompts).toContain(
-      'Summarize governed remediation risks',
+      'Review pending approvals and safest next step',
     );
     expect(JSON.stringify(handoff)).not.toContain('systemctl restart workload.service');
   });

frontend-modern/src/features/patrol/patrolInvestigationContextModel.ts

@@ -410,6 +410,8 @@ function buildPatrolAssessmentAssistantBriefing(
   const recentChanges = normalizeAssessmentRecentChanges(input.supportingEvidence?.recentChanges);
   const correlations = normalizeAssessmentCorrelations(input.supportingEvidence?.correlations);
   const findingEvidence = findings.map(formatAssessmentFindingEvidence).filter(isNonEmptyString);
+  const handoffActions = buildPatrolAssessmentHandoffActions(input);
+  const actionPosture = buildPatrolAssessmentActionPosture(input, handoffActions);
   const supportingEvidence = [
     ...recentChanges.map(formatAssessmentRecentChangeEvidence),
     ...correlations.map(formatAssessmentCorrelationEvidence),
@@ -426,8 +428,8 @@ function buildPatrolAssessmentAssistantBriefing(
       .filter(isNonEmptyString)
       .slice(0, 4),
     evidence: [...findingEvidence.slice(0, 3), ...supportingEvidence].slice(0, 5),
-    actionLabel: 'Discuss Patrol assessment',
-    safetyNote: 'Diagnostics and remediation require governed approval.',
+    actionLabel: actionPosture.actionLabel,
+    safetyNote: actionPosture.safetyNote,
     suggestedPrompts: buildPatrolAssessmentSuggestedPrompts(input, {
       findings,
       recentChanges,
@@ -436,6 +438,76 @@ function buildPatrolAssessmentAssistantBriefing(
   };
 }
 
+function buildPatrolAssessmentActionPosture(
+  input: PatrolAssessmentAssistantHandoffInput,
+  handoffActions: AIChatHandoffAction[],
+): { actionLabel: string; safetyNote: string } {
+  const pendingApprovalCount = normalizeAssessmentPendingApprovalCount(input.activeFindings);
+  const actionCount = handoffActions.length;
+  const hasDryRunPosture = handoffActions.some((action) =>
+    Boolean(normalizeText(action.actionDryRunSummary) || normalizeText(action.actionPreflight)),
+  );
+  const hasDestructiveAction = handoffActions.some((action) => Boolean(action.destructive));
+  const hasApprovalPolicy = handoffActions.some((action) =>
+    Boolean(normalizeText(action.actionApprovalPolicy)),
+  );
+
+  if (pendingApprovalCount > 0) {
+    return {
+      actionLabel: `${formatAssessmentMetricCount(
+        'Pending governed approvals',
+        pendingApprovalCount,
+      )} attached`,
+      safetyNote: formatAssessmentActionSafetyNote({
+        primary: 'Review approvals in the governed flow',
+        hasDryRunPosture,
+        hasDestructiveAction,
+        hasApprovalPolicy,
+      }),
+    };
+  }
+
+  if (actionCount > 0) {
+    return {
+      actionLabel: `${formatAssessmentMetricCount(
+        'Governed action references',
+        actionCount,
+      )} attached`,
+      safetyNote: formatAssessmentActionSafetyNote({
+        primary: 'Review action posture in the governed flow',
+        hasDryRunPosture,
+        hasDestructiveAction,
+        hasApprovalPolicy,
+      }),
+    };
+  }
+
+  return {
+    actionLabel: 'Discuss Patrol assessment',
+    safetyNote: 'Diagnostics and remediation require governed approval.',
+  };
+}
+
+function formatAssessmentActionSafetyNote(input: {
+  primary: string;
+  hasDryRunPosture: boolean;
+  hasDestructiveAction: boolean;
+  hasApprovalPolicy: boolean;
+}): string {
+  const parts = [input.primary];
+  if (input.hasApprovalPolicy) {
+    parts.push('approval policy is attached');
+  }
+  if (input.hasDryRunPosture) {
+    parts.push('dry-run posture is attached');
+  }
+  if (input.hasDestructiveAction) {
+    parts.push('destructive actions remain approval-bound');
+  }
+  parts.push('raw command payloads stay out of Assistant');
+  return `${parts.join('; ')}.`;
+}
+
 function buildPatrolAssessmentSuggestedPrompts(
   input: PatrolAssessmentAssistantHandoffInput,
   normalized: {
@@ -461,7 +533,11 @@ function buildPatrolAssessmentSuggestedPrompts(
   }
 
   if (hasGovernedAction) {
-    prompts.push('Summarize governed remediation risks');
+    prompts.push(
+      normalizeAssessmentPendingApprovalCount(input.activeFindings) > 0
+        ? 'Review pending approvals and safest next step'
+        : 'Summarize governed remediation risks',
+    );
   } else if (activeFindingCount > 0) {
     prompts.push('List evidence to verify before action');
   } else if (hasSupportingEvidence) {