Security Case Management

Case Management

Manage SOC cases through the Kibana Cases API. All cases are scoped to securitySolution — this skill operates exclusively within Elastic Security. Cases appear in Kibana Security and can be assigned to analysts, linked to alerts, and pushed to external incident management systems via connectors.

Prerequisites

Install dependencies before first use from the skills/security directory:

cd skills/security && npm install

Set the required environment variables (or add them to a .env file in the workspace root):

export KIBANA_URL="https://your-cluster.kb.cloud.example.com:443"
export KIBANA_API_KEY="your-kibana-api-key"

When to use

• Creating a case after alert triage (classification, IOCs, findings)
• Searching for existing cases to correlate related alerts
• Adding investigation comments or attaching alerts to an existing case
• Updating case status or severity
• Listing recent cases for review

When NOT to use

• Do not use this skill for Observability or Elasticsearch cases — it hardcodes owner: securitySolution
• Do not use for cases outside the Security solution space

Execution rules

• Start executing tools immediately — do not read SKILL.md, browse the workspace, or list files first.
• Report tool output faithfully. Copy case IDs, titles, tags, severities, and counts exactly as returned by the API. Do not abbreviate case IDs, truncate titles, invent details, or round numbers.
• When the API returns zero results, state that explicitly — do not guess at possible results.
• When listing or finding cases, report the exact total count from the API response and present each case with its verbatim title, severity, and status.

Quick start

All commands run from the workspace root. All output is JSON. Call the tools directly — do not read the skill file or explore the workspace first. For attach-alert/attach-alerts, --rule-id and --rule-name are required by the Kibana API (use --rule-id unknown --rule-name unknown if unknown). Use attach-alerts for batch with automatic rate-limit retry and 2-second spacing between API calls.

Common multi-step workflows

Finding cases for a host: Use find --search "<hostname>" to search by hostname across title, description, and comments. Alternatively use find --tags "agent_id:<agent_id>" if the agent ID is known. Always add --status open to filter to active cases only. Report the exact total count and each case title verbatim from the API response.

# Create (syncAlerts enabled by default; disable with --sync-alerts false)
node skills/security/case-management/scripts/case-manager.js create --title "Malicious DLL sideloading on host1" --description "Crypto clipper malware detected via DLL sideloading..." --tags "classification:malicious" "confidence:88" "mitre:T1574.002" --severity critical --yes

# Find, list, get
node skills/security/case-management/scripts/case-manager.js find --tags "agent_id:550888e5-357d-4bc1-a154-486eb7b4e076"
node skills/security/case-management/scripts/case-manager.js find --search "DLL sideloading" --status open
node skills/security/case-management/scripts/case-manager.js list --status open --per-page 10
node skills/security/case-management/scripts/case-manager.js get --case-id <case_id>

# Attach single alert
node skills/security/case-management/scripts/case-manager.js attach-alert --case-id <case_id> --alert-id <alert_doc_id> --alert-index .ds-.alerts-security.alerts-default-2025.12.01-000013 --rule-id <rule_uuid> --rule-name "Malware Detection Alert"

# Attach multiple alerts (batch)
node skills/security/case-management/scripts/case-manager.js attach-alerts --case-id <case_id> --alert-ids <id1> <id2> <id3> --alert-index .ds-.alerts-security.alerts-default-2026.02.16-000016 --rule-id <rule_uuid> --rule-name "Malware Detection Alert"

# Add comment, update (--tags merges with existing tags, does not replace)
node skills/security/case-management/scripts/case-manager.js add-comment --case-id <case_id> --comment "Process tree analysis shows..."
node skills/security/case-management/scripts/case-manager.js update --case-id <case_id> --status closed --severity low --yes

Write operations (create, update) prompt for confirmation by default. Pass --yes to skip the prompt (required when called by an agent).

Reporting list and find results

When reporting results from list or find:

1. State the exact total count from the JSON response (e.g., "There are 12 open cases total").
2. Present each case as a compact one-line entry: <title> | <severity> | <case_id_short> | <created_at>. Copy the exact title verbatim from the title field — do not rephrase, abbreviate, or summarize.
3. If the user asked for N cases, present exactly N entries (or fewer if fewer exist). Do not add extra columns (alerts count, description, status) unless the user specifically requested them.
4. Do not add information beyond what the API returned. If a field is null or missing, omit it.
5. After presenting the results, stop. Do not add analysis, commentary, or observations about the cases.

Tag conventions

Use structured tags for machine-searchable metadata:

Case severity mapping

Known limitations

syncAlerts is security-only

The syncAlerts setting (enabled by default) synchronizes case status with attached alert statuses. This feature is only available for Security Solution cases. Pass --sync-alerts false when creating a case if alert sync is not needed.

Rate limiting

The Kibana API enforces rate limits. When attaching multiple alerts, the attach-alerts batch command automatically handles 429 responses with retry. If using attach-alert one at a time, space calls ~10 seconds apart.

find --search on Serverless

The find --search parameter may return 500 errors on Kibana Serverless deployments. Use find --tags for filtering instead, or list to browse recent cases.