spine-framework-cortex 0.2.21 → 0.2.22

package/pages/support/TicketDetailPage.tsx

@@ -1,4 +1,4 @@
-import { useEffect, useState } from 'react'
+import { useEffect, useRef, useState } from 'react'
 import { useParams, useNavigate } from 'react-router-dom'
 import { apiFetch } from '@core/lib/api'
 import { Button } from '@core/components/ui/button'
@@ -7,7 +7,8 @@ import { Badge } from '@core/components/ui/badge'
 import { Skeleton } from '@core/components/ui/skeleton'
 import { ScrollArea } from '@core/components/ui/scroll-area'
 import { Tabs, TabsList, TabsTrigger, TabsContent } from '@core/components/ui/tabs'
-import { ArrowLeft, Send, Lock, Globe, Bot, CheckCircle, AlertCircle, FileText, Building2, CreditCard, Package, BookOpen, Eye, EyeOff, Brain, Clock, TrendingUp, Tag as TagIcon } from 'lucide-react'
+import { Textarea } from '@core/components/ui/textarea'
+import { ArrowLeft, Send, Lock, Globe, Bot, CheckCircle, AlertCircle, FileText, Building2, CreditCard, Package, BookOpen, Eye, EyeOff, Brain, Clock, TrendingUp, Tag as TagIcon, Sparkles } from 'lucide-react'
 import { useAuth } from '@core/contexts/AuthContext'
 
 interface Ticket { 
@@ -367,145 +368,143 @@ function CaseDataPanel({ ticket }: { ticket: Ticket | null }) {
   )
 }
 
-// AI Metadata Panel
-function AIMetadataPanel({ ticket, onGenerateKB }: { ticket: Ticket | null; onGenerateKB?: () => void }) {
-  if (!ticket?.data?.aim_confidence_threshold && !ticket?.data?.aim_confidence_at_response && !ticket?.data?.aim_escalation_reason) {
-    return (
-      <div className="h-full flex flex-col">
-        <div className="px-4 py-2 border-b border-border bg-purple-50">
-          <p className="text-xs font-medium uppercase tracking-wide text-purple-700">AI Analysis</p>
-        </div>
-        <div className="flex-1 p-4 flex items-center justify-center text-sm text-muted-foreground">
-          No AI metadata available for this ticket.
-        </div>
-      </div>
-    )
-  }
+// Solution AI Panel — internal AI collaborator for support agents
+interface AIMessage { role: 'user' | 'ai'; content: string }
 
-  const ai = ticket.data
-  const isResolved = ticket.status === 'resolved' || ticket.status === 'closed'
+function SolutionAIPanel({ ticket, onGenerateKB }: { ticket: Ticket | null; onGenerateKB?: () => void }) {
+  const [messages, setMessages] = useState<AIMessage[]>([])
+  const [input, setInput] = useState('')
+  const [sending, setSending] = useState(false)
+  const [threadId, setThreadId] = useState<string | null>(null)
+  const bottomRef = useRef<HTMLDivElement>(null)
+  const ai = ticket?.data
+  const isResolved = ticket?.status === 'resolved' || ticket?.status === 'closed'
+
+  // Auto-scroll to bottom when new messages arrive
+  useEffect(() => {
+    bottomRef.current?.scrollIntoView({ behavior: 'smooth' })
+  }, [messages])
+
+  const sendMessage = async () => {
+    if (!input.trim() || !ticket?.id || sending) return
+    const userText = input.trim()
+    setInput('')
+    setMessages(prev => [...prev, { role: 'user', content: userText }])
+    setSending(true)
+    try {
+      const isFirst = !threadId
+      const res = await apiFetch(
+        `/.netlify/functions/custom_support-solution?action=${isFirst ? 'start' : 'message'}`,
+        {
+          method: 'POST',
+          headers: { 'Content-Type': 'application/json' },
+          body: JSON.stringify(isFirst
+            ? { ticket_id: ticket.id, message: userText }
+            : { thread_id: threadId, message: userText }
+          ),
+        }
+      ).then(r => r.json())
+
+      if (res.threadId && !threadId) setThreadId(res.threadId)
+      if (res.content) setMessages(prev => [...prev, { role: 'ai', content: res.content }])
+    } catch (err) {
+      setMessages(prev => [...prev, { role: 'ai', content: 'Error contacting Solution AI. Please try again.' }])
+    } finally {
+      setSending(false)
+    }
+  }
 
   return (
     <div className="h-full flex flex-col">
-      <div className="px-4 py-2 border-b border-border bg-purple-50">
-        <p className="text-xs font-medium uppercase tracking-wide text-purple-700">AI Analysis</p>
+      {/* Header */}
+      <div className="px-4 py-2 border-b border-border bg-purple-50 shrink-0">
+        <div className="flex items-center justify-between">
+          <div className="flex items-center gap-1.5">
+            <Sparkles className="h-3.5 w-3.5 text-purple-600" />
+            <p className="text-xs font-medium uppercase tracking-wide text-purple-700">Solution AI</p>
+          </div>
+          <p className="text-xs text-purple-500">Internal only</p>
+        </div>
       </div>
-      <ScrollArea className="flex-1 p-4">
-        <div className="space-y-4">
-          {/* Confidence Score */}
+
+      {/* Triage metadata strip */}
+      {(ai?.aim_confidence_at_response || ai?.aim_escalation_reason) && (
+        <div className="px-3 py-1.5 border-b border-border bg-muted/40 shrink-0 flex items-center gap-3 text-xs text-muted-foreground">
           {ai.aim_confidence_at_response && (
-            <div className="space-y-1">
-              <div className="flex items-center justify-between text-sm">
-                <span className="text-muted-foreground">Confidence</span>
-                <span className={`font-mono font-medium ${ai.aim_confidence_at_response >= 0.75 ? 'text-green-600' : 'text-amber-600'}`}>
-                  {Math.round(ai.aim_confidence_at_response * 100)}%
-                </span>
-              </div>
-              {ai.aim_confidence_threshold && (
-                <p className="text-xs text-muted-foreground">Threshold: {Math.round(ai.aim_confidence_threshold * 100)}%</p>
-              )}
-            </div>
+            <span className={ai.aim_confidence_at_response >= 0.75 ? 'text-green-600' : 'text-amber-600'}>
+              Triage confidence: {Math.round(ai.aim_confidence_at_response * 100)}%
+            </span>
           )}
-
           {ai.aim_escalation_reason && ai.aim_escalation_reason !== 'none' && (
-            <div className="flex items-start gap-2 p-2 bg-amber-50 rounded-lg">
-              <AlertCircle className="h-4 w-4 text-amber-600 mt-0.5" />
-              <div className="text-sm">
-                <p className="font-medium text-amber-700">Escalated</p>
-                <p className="text-amber-600 text-xs">{ai.aim_escalation_reason.replace('_', ' ')}</p>
-              </div>
-            </div>
-          )}
-
-          {/* Problem Statement */}
-          {ai.aim_problem_statement && (
-            <div className="space-y-1">
-              <p className="text-xs font-medium text-muted-foreground uppercase">Problem</p>
-              <p className="text-sm bg-muted p-2 rounded">{ai.aim_problem_statement}</p>
-            </div>
+            <span className="text-amber-600 flex items-center gap-1">
+              <AlertCircle className="h-3 w-3" /> Escalated
+            </span>
           )}
-
-          {/* Solution Path */}
-          {ai.aim_solution_path && (
-            <div className="space-y-1">
-              <p className="text-xs font-medium text-muted-foreground uppercase">Solution</p>
-              <p className="text-sm bg-muted p-2 rounded">{ai.aim_solution_path}</p>
-            </div>
-          )}
-
-          {/* Tools Used */}
-          {ai.aim_tools_used && ai.aim_tools_used.length > 0 && (
-            <div className="space-y-1">
-              <p className="text-xs font-medium text-muted-foreground uppercase">Tools Used</p>
-              <div className="flex flex-wrap gap-1">
-                {ai.aim_tools_used.map((tool, i) => (
-                  <Badge key={i} variant="outline" className="text-xs">{tool}</Badge>
-                ))}
+        </div>
+      )}
+
+      {/* Chat history */}
+      <ScrollArea className="flex-1 px-3 py-2">
+        {messages.length === 0 ? (
+          <div className="flex flex-col items-center justify-center h-32 text-center text-muted-foreground text-xs gap-2">
+            <Bot className="h-6 w-6" />
+            <p>Ask Solution AI for help with this case.</p>
+            <p className="text-xs opacity-70">Searches KB · recalls similar cases · drafts replies</p>
+          </div>
+        ) : (
+          <div className="space-y-2">
+            {messages.map((msg, i) => (
+              <div key={i} className={`flex ${msg.role === 'user' ? 'justify-end' : 'justify-start'}`}>
+                <div className={`max-w-[90%] rounded-lg px-3 py-2 text-xs whitespace-pre-wrap ${
+                  msg.role === 'user'
+                    ? 'bg-primary text-primary-foreground'
+                    : 'bg-muted text-foreground'
+                }`}>
+                  {msg.content}
+                </div>
               </div>
-            </div>
-          )}
-
-          {/* Post-mortem Section */}
-          {isResolved && (
-            <>
-              <Separator />
-              <div className="space-y-3">
-                <p className="text-xs font-medium text-muted-foreground uppercase">Post-mortem</p>
-
-                {postmortem?.root_cause_category && (
-                  <p className="text-sm">
-                    <span className="text-muted-foreground">Root cause:</span>{' '}
-                    {postmortem.root_cause_category}
-                  </p>
-                )}
-
-                {postmortem?.resolution_time_minutes && (
-                  <p className="text-sm">
-                    <span className="text-muted-foreground">Resolution time:</span>{' '}
-                    {Math.round(postmortem.resolution_time_minutes)} min
-                  </p>
-                )}
-
-                {postmortem?.customer_satisfaction !== undefined && (
-                  <div className="flex items-center gap-2">
-                    <span className="text-sm text-muted-foreground">Satisfaction:</span>
-                    <div className="flex">
-                      {[1, 2, 3, 4, 5].map((star) => (
-                        <span 
-                          key={star} 
-                          className={`text-lg ${star <= (postmortem.customer_satisfaction || 0) ? 'text-amber-400' : 'text-gray-300'}`}
-                        >
-                          ★
-                        </span>
-                      ))}
-                    </div>
-                  </div>
-                )}
-
-                {/* KB Generation */}
-                <div className="pt-2">
-                  {postmortem?.kb_generated ? (
-                    <div className="flex items-center gap-2 text-green-600 text-sm">
-                      <CheckCircle className="h-4 w-4" />
-                      KB article created
-                    </div>
-                  ) : (
-                    <Button 
-                      variant="outline" 
-                      size="sm" 
-                      className="w-full gap-1 text-xs"
-                      onClick={onGenerateKB}
-                    >
-                      <BookOpen className="h-3 w-3" />
-                      Generate KB Article
-                    </Button>
-                  )}
+            ))}
+            {sending && (
+              <div className="flex justify-start">
+                <div className="bg-muted rounded-lg px-3 py-2 text-xs text-muted-foreground animate-pulse">
+                  Thinking…
                 </div>
               </div>
-            </>
+            )}
+            <div ref={bottomRef} />
+          </div>
+        )}
+      </ScrollArea>
+
+      {/* KB generation (resolved only) */}
+      {isResolved && (
+        <div className="px-3 py-2 border-t border-border shrink-0">
+          {ai?.kb_proposed_kb_id ? (
+            <div className="flex items-center gap-2 text-green-600 text-xs">
+              <CheckCircle className="h-3.5 w-3.5" /> KB article created
+            </div>
+          ) : (
+            <Button variant="outline" size="sm" className="w-full gap-1 text-xs" onClick={onGenerateKB}>
+              <BookOpen className="h-3 w-3" /> Generate KB Article
+            </Button>
           )}
         </div>
-      </ScrollArea>
+      )}
+
+      {/* Input */}
+      <div className="px-3 py-2 border-t border-border shrink-0 flex gap-2">
+        <Textarea
+          value={input}
+          onChange={e => setInput(e.target.value)}
+          onKeyDown={e => { if (e.key === 'Enter' && !e.shiftKey) { e.preventDefault(); sendMessage() } }}
+          placeholder="Ask Solution AI…"
+          className="flex-1 text-xs resize-none min-h-[36px] max-h-[100px]"
+          rows={1}
+        />
+        <Button size="sm" onClick={sendMessage} disabled={!input.trim() || sending} className="shrink-0">
+          <Send className="h-3.5 w-3.5" />
+        </Button>
+      </div>
     </div>
   )
 }
@@ -838,7 +837,6 @@ export default function TicketDetailPage() {
   }
 
   const handleGenerateKB = () => {
-    // Trigger KB generation flow - would navigate to redaction review
     navigate(`/cortex/support/${id}/kb-review`)
   }
 
@@ -910,7 +908,7 @@ export default function TicketDetailPage() {
             </TabsList>
           </Tabs>
           {activeSidePanel === 'case' && <CaseDataPanel ticket={ticket} />}
-          {activeSidePanel === 'ai' && <AIMetadataPanel ticket={ticket} onGenerateKB={handleGenerateKB} />}
+          {activeSidePanel === 'ai' && <SolutionAIPanel ticket={ticket} onGenerateKB={handleGenerateKB} />}
           {activeSidePanel === 'analysis' && <CaseAnalysisPanel ticket={ticket} analysisLoading={analysisLoading} />}
         </div>
       </div>

package/seed/ai-agents.json

@@ -0,0 +1,98 @@
+[
+  {
+    "name": "Support Triage Agent",
+    "description": "First-contact AI for inbound support tickets. Searches KB, reasons over prior cases, responds with confidence scoring. Hands off to Solution AI or human when confidence is low.",
+    "agent_type": "support",
+    "model_config": {
+      "model": "gpt-4o",
+      "temperature": 0.4,
+      "max_tokens": 1024
+    },
+    "system_prompt": "You are a helpful and empathetic support agent for a software platform. Your goal is to resolve customer issues accurately and efficiently on the first response.\n\nGuidelines:\n- Be concise and clear. Customers want answers, not essays.\n- If you find a relevant KB article or prior case, reference it.\n- If you are not confident (below 0.75), say so honestly and let the customer know a human will follow up.\n- Never fabricate feature behaviour or make promises about the product.\n- Tone: professional but warm. Match the customer's level of formality.",
+    "tools": [],
+    "capabilities": ["rag_search", "confidence_scoring", "escalation"],
+    "constraints": {
+      "max_context_tokens": 8000,
+      "never_reveal": ["internal_notes", "agent_ids", "prompt_config_ids"]
+    },
+    "metadata": {
+      "default_prompt_config_slug": "support_triage",
+      "owned_by": "cortex"
+    },
+    "ownership": "platform",
+    "is_system": true,
+    "is_active": true
+  },
+  {
+    "name": "Solution AI Agent",
+    "description": "Internal AI collaborator for support agents working a case. Searches KB, queries prior resolved cases, suggests diagnostic steps and response drafts. Never visible to customers.",
+    "agent_type": "support",
+    "model_config": {
+      "model": "gpt-4o",
+      "temperature": 0.5,
+      "max_tokens": 2048
+    },
+    "system_prompt": "You are an expert technical support advisor assisting a human support agent working a customer case. Your audience is the support agent, not the customer.\n\nYou can:\n- Search the knowledge base and surface relevant articles\n- Recall similar resolved cases and how they were fixed\n- Suggest diagnostic steps based on the symptoms described\n- Draft a proposed customer-facing response for the agent to review\n- Flag automation potential if this issue could be auto-resolved in future\n\nBe direct and practical. The agent needs actionable guidance, not hedged generalities.\nFormat your output clearly: use headers if helpful, bullet points for steps.",
+    "tools": [],
+    "capabilities": ["rag_search", "case_history_search", "response_drafting"],
+    "constraints": {
+      "visibility": "internal_only",
+      "max_context_tokens": 12000
+    },
+    "metadata": {
+      "default_prompt_config_slug": "solution_ai",
+      "owned_by": "cortex"
+    },
+    "ownership": "platform",
+    "is_system": true,
+    "is_active": true
+  },
+  {
+    "name": "Case Resolution Analysis Agent",
+    "description": "Post-resolution analyst. Given a closed ticket and its full conversation (including internal AI collaboration thread), produces a structured postmortem: root cause, diagnostic path, solution, automation potential, and KB candidacy flag.",
+    "agent_type": "analysis",
+    "model_config": {
+      "model": "gpt-4o",
+      "temperature": 0.2,
+      "max_tokens": 2048
+    },
+    "system_prompt": "You are a case analysis specialist. When given a resolved support ticket — including the customer conversation, the internal support-agent collaboration thread, and any escalation notes — you produce a structured postmortem.\n\nYour output must be a JSON object with these exact fields:\n- reported_issue: what the customer described (their words)\n- true_problem: the actual root cause (may differ significantly from how it was reported)\n- diagnostic_steps: ordered array of steps taken to identify the root cause\n- solution_steps: ordered array of steps taken to resolve it\n- final_solution: one-sentence concise answer\n- customer_temperature: 'positive' | 'neutral' | 'frustrated' | 'escalated'\n- time_to_resolution: estimated minutes from first message to resolution\n- escalation_required: boolean\n- back_and_forth_count: integer — number of customer messages\n- sentiment_progression: array of sentiment labels across the conversation\n- automation_potential: 'high' | 'medium' | 'low' — could this be auto-resolved next time?\n- kb_candidate: boolean — would a KB article have prevented this ticket?\n- suggested_tags: array of tag objects { slug, name, purpose, category, applicable_to }\n- confidence_score: 0.0-1.0\n- analysis_summary: 2-3 sentence plain English summary for the ticket record",
+    "tools": [],
+    "capabilities": ["structured_output", "sentiment_analysis", "case_classification"],
+    "constraints": {
+      "output_format": "json",
+      "max_context_tokens": 16000
+    },
+    "metadata": {
+      "default_prompt_config_slug": "case_analysis_prompt",
+      "owned_by": "cortex"
+    },
+    "ownership": "platform",
+    "is_system": true,
+    "is_active": true
+  },
+  {
+    "name": "KB Redaction Agent",
+    "description": "Prepares resolved cases for publication as KB articles. Identifies PII, account-specific details, internal references, and confidential data. Produces redaction suggestions and a generalised, customer-facing rewrite.",
+    "agent_type": "content",
+    "model_config": {
+      "model": "gpt-4o",
+      "temperature": 0.1,
+      "max_tokens": 3000
+    },
+    "system_prompt": "You are a knowledge base editorial agent. You receive the raw text of a resolved support case and must prepare it for public publication as a KB article.\n\nYour tasks:\n1. IDENTIFY all text that should be redacted before publishing. Categories:\n   - pii: names, emails, phone numbers, addresses, usernames\n   - account_specific: account names, account IDs, organisation-specific data\n   - confidential: internal pricing, unreleased features, internal process details\n   - internal_reference: ticket IDs, internal tool names, employee names\n\n2. GENERALISE the solution: rewrite the content to be universally applicable, removing all account-specific context while preserving the diagnostic and solution logic.\n\nOutput must be a JSON object:\n{\n  \"original_content\": \"<the input text unchanged>\",\n  \"redacted_content\": \"<fully redacted and generalised version>\",\n  \"suggestions\": [\n    {\n      \"id\": \"<uuid>\",\n      \"start_index\": <int>,\n      \"end_index\": <int>,\n      \"original_text\": \"<exact text to redact>\",\n      \"redacted_text\": \"<replacement text, e.g. [Account Name]>\",\n      \"sensitivity_level\": \"high|medium|low\",\n      \"reasoning\": \"<one sentence why>\",\n      \"category\": \"pii|confidential|account_specific|internal_reference\"\n    }\n  ],\n  \"confidence_score\": 0.0-1.0,\n  \"processing_metadata\": { \"model_used\": \"<model>\", \"temperature\": 0.1, \"tokens_consumed\": <int> }\n}",
+    "tools": [],
+    "capabilities": ["structured_output", "redaction", "content_generalisation"],
+    "constraints": {
+      "output_format": "json",
+      "max_context_tokens": 12000
+    },
+    "metadata": {
+      "default_prompt_config_slug": "kb_generator",
+      "owned_by": "cortex"
+    },
+    "ownership": "platform",
+    "is_system": true,
+    "is_active": true
+  }
+]

package/seed/pipelines.json

@@ -0,0 +1,29 @@
+[
+  {
+    "name": "support_escalation",
+    "description": "Fired by agent-runner when triage confidence falls below threshold. Marks the thread as pending human review and notifies the support team.",
+    "trigger_type": "event",
+    "config": {
+      "event": "agent.inference.low_confidence",
+      "scope": "account"
+    },
+    "stages": [
+      {
+        "stage_type": "send_notification",
+        "config": {
+          "message": "A support ticket has been escalated to human review — confidence too low for AI response.",
+          "entity_type": "threads",
+          "entity_id_field": "thread_id",
+          "recipients": ["role:support_admin", "role:support"]
+        },
+        "continue_on_error": true
+      }
+    ],
+    "metadata": {
+      "note": "Thread escalation_status is already set to 'pending' by agent-runner before this pipeline fires. Ticket status is updated by the calling function (custom_support-triage) from agentMsg.data.escalated."
+    },
+    "ownership": "platform",
+    "is_system": true,
+    "is_active": true
+  }
+]

package/seed/prompt-configs.json

@@ -0,0 +1,84 @@
+[
+  {
+    "name": "Support Triage",
+    "slug": "support_triage",
+    "context_template": "A customer has submitted the following support request:\n\n{{user_message}}\n\nReview the knowledge base articles and prior case history provided in your context. Respond helpfully and accurately. If you are not confident in your answer (below 0.75), set escalate=true.",
+    "is_multi_turn": true,
+    "max_history_messages": 20,
+    "confidence_threshold": 0.75,
+    "escalation_action": "pipeline",
+    "escalation_target": "support_escalation",
+    "output_mode": "json",
+    "knowledge_sources": [],
+    "available_tools": [],
+    "metadata": {
+      "knowledge_scope": ["kb_article", "resolved_case"],
+      "note": "knowledge_sources is resolved at runtime to the requesting account's KB + platform KB. Empty array = full scope search."
+    },
+    "ownership": "platform",
+    "is_system": true,
+    "is_active": true
+  },
+  {
+    "name": "Solution AI",
+    "slug": "solution_ai",
+    "context_template": "A support agent is working on the following case and needs your assistance:\n\nTicket: {{ticket_title}}\nCustomer issue: {{ticket_description}}\n\nConversation so far:\n{{conversation_history}}\n\nAgent question: {{user_message}}\n\nProvide practical, actionable guidance. If you can draft a suggested customer-facing response, include it under 'Suggested Response:'.",
+    "is_multi_turn": true,
+    "max_history_messages": 30,
+    "confidence_threshold": 0.0,
+    "escalation_action": null,
+    "escalation_target": null,
+    "output_mode": "text",
+    "knowledge_sources": [],
+    "available_tools": [],
+    "metadata": {
+      "visibility": "internal",
+      "knowledge_scope": ["kb_article", "resolved_case"],
+      "note": "No escalation — this agent assists human agents. confidence_threshold=0 disables auto-escalation."
+    },
+    "ownership": "platform",
+    "is_system": true,
+    "is_active": true
+  },
+  {
+    "name": "Case Analysis Prompt",
+    "slug": "case_analysis_prompt",
+    "context_template": "Analyze this resolved support ticket:\n\nTicket: {{ticket_title}}\nDescription: {{ticket_description}}\nCreated: {{created_at}}\nResolved: {{resolved_at}}\nStatus: {{status}}\nPriority: {{priority}}\n\nCustomer Conversation:\n{{conversation_history}}\n\nInternal Support Thread:\n{{internal_thread}}\n\nProduce a complete structured postmortem in the JSON format specified in your system prompt.",
+    "is_multi_turn": false,
+    "max_history_messages": 0,
+    "confidence_threshold": 0.0,
+    "escalation_action": null,
+    "escalation_target": null,
+    "output_mode": "json",
+    "knowledge_sources": [],
+    "available_tools": [],
+    "metadata": {
+      "requires_resolved_ticket": true,
+      "includes_internal_thread": true
+    },
+    "ownership": "platform",
+    "is_system": true,
+    "is_active": true
+  },
+  {
+    "name": "KB Generator",
+    "slug": "kb_generator",
+    "context_template": "Prepare the following resolved case content for KB publication:\n\n{{content}}\n\nIdentify all redactable text and produce a generalised, customer-facing version suitable for publishing as a public knowledge base article.",
+    "is_multi_turn": false,
+    "max_history_messages": 0,
+    "confidence_threshold": 0.0,
+    "escalation_action": null,
+    "escalation_target": null,
+    "output_mode": "json",
+    "requires_review": true,
+    "knowledge_sources": [],
+    "available_tools": [],
+    "metadata": {
+      "redaction_categories": ["pii", "confidential", "account_specific", "internal_reference"],
+      "auto_accept_sensitivity": ["high"]
+    },
+    "ownership": "platform",
+    "is_system": true,
+    "is_active": true
+  }
+]