Splunk Enterprise Security Config

Prerequisites

Tool or access	Purpose	Verify
Bash and Python 3	Run bundled setup and validation helpers	bash --version && python3 --version
Required product/platform access	Inspect or configure the selected target	Complete the documented preflight
Credential files for live modes	Keep secrets out of chat	Verify paths only

Workflow Overview

┌───────────┐   ┌───────────────┐   ┌───────────────┐   ┌─────────────────┐
│ Preflight │ → │ Render/review │ → │ Apply/handoff │ → │ Validate evidence │
└───────────┘   └───────────────┘   └───────────────┘   └─────────────────┘

When to Activate

• Configure, tune, validate, or operationalize Splunk Enterprise Security.
• Preview and review the splunk enterprise security config workflow before any live apply phase.
• Diagnose failed prerequisites, generated assets, configuration, or validation evidence.

Scope

Follow the documented read-only or render-first path whenever it is available. This skill does not imply permission to mutate live systems. Require explicit apply flags, protected credentials, and operator review for state changes.

Examples

Inspect the supported setup modes before selecting one:

bash skills/splunk-enterprise-security-config/scripts/setup.sh --help

Expected output: usage, supported modes, and required arguments are displayed without changing the target environment.

Inspect validation modes before running completion checks:

bash skills/splunk-enterprise-security-config/scripts/validate.sh --help

Expected output: offline, live, and completion options are displayed when the skill supports them; help exits without mutation.

Troubleshooting

Issue	Cause	Resolution
Preflight fails	A required tool or access path is missing	Resolve it before rendering or applying
Rendered assets are incomplete	Required non-secret inputs are absent	Complete intake and render again
Apply is blocked	Review, credentials, or explicit acceptance is missing	Use the documented handoff
Validation is incomplete	Live evidence is unavailable	Record the gap and keep completion open

Configures and validates Splunk Enterprise Security after SplunkEnterpriseSecuritySuite has been installed and post-install setup has run.

Agent Behavior

Never ask for secrets in chat. Splunk credentials are read from the project-root credentials file, falling back to ~/.splunk/credentials. If neither exists, guide the user to run:

bash skills/shared/scripts/setup_credentials.sh

Credential-bearing REST requests require a credential-free HTTPS origin and never follow redirects. Plaintext HTTP is refused unless an operator explicitly sets SPLUNK_ALLOW_INSECURE_HTTP=true for an isolated, short-lived lab; SPLUNK_VERIFY_SSL=false does not authorize HTTP.

Before applying changes, identify the deployment shape:

• Standalone search head
• Dedicated distributed search head
• Search head cluster deployer
• Splunk Cloud Platform
• Hybrid on-prem ES search head searching remote indexers

Use this skill for ES configuration. Use splunk-enterprise-security-install first if ES is not installed.

Quick Start

Validate current ES configuration:

bash skills/splunk-enterprise-security-config/scripts/validate.sh

Preview a full declarative ES configuration plan:

bash skills/splunk-enterprise-security-config/scripts/setup.sh \
  --spec skills/splunk-enterprise-security-config/templates/es-config.example.yaml

Apply an explicit ES configuration spec:

bash skills/splunk-enterprise-security-config/scripts/setup.sh \
  --spec es-config.yaml \
  --apply \
  --validate

Apply low-risk baseline settings on an on-prem search tier:

bash skills/splunk-enterprise-security-config/scripts/setup.sh \
  --set-lookup-order \
  --set-managed-roles ess_analyst,ess_user \
  --validate

Create ES indexes on the target index tier or standalone instance:

bash skills/splunk-enterprise-security-config/scripts/setup.sh \
  --create-core-indexes \
  --create-mission-control-indexes \
  --create-exposure-indexes \
  --max-size-mb 512000 \
  --validate

Configuration Workflow

1. Read the reference:

   skills/splunk-enterprise-security-config/reference.md
   

2. Validate install state with the install skill:

   bash skills/splunk-enterprise-security-install/scripts/validate.sh
   

3. Deploy technology add-ons:

   • Keep ES framework apps (SA- and DA-ESS-) enabled.
   • Deploy data-source TAs to forwarders and indexers based on each TA's docs.
   • For clustered on-prem indexers, generate and redeploy Splunk_TA_ForIndexers after each ES release or TA change.

4. Configure indexes and role search scope:

   • Create ES and TA indexes on the index tier for distributed deployments.
   • Add security data indexes to ess_user, ess_analyst, and admin/sc_admin search scope. Do not include summary indexes as default searched indexes.

5. Configure CIM and data models:

   • Constrain CIM data models to relevant indexes for performance.
   • Review acceleration storage and retention.
   • Keep ES acceleration enforcement enabled for required models unless there is an explicit performance plan.

6. Configure SOC enrichment:

   • Assets and identities
   • Native threat intelligence or Threat Intelligence Management (Cloud)
   • Detections, finding-based detections, risk-based alerting, and MITRE annotations
   • Analyst queue/Mission Control queues and response workflows
   • SOAR, Attack Analyzer, UEBA, exposure analytics, and Splunk Cloud Connect where licensed and supported
   • ES AI Assistant settings and Federated Analytics / ASL readiness where licensed and supported

7. Validate and iterate:

   bash skills/splunk-enterprise-security-config/scripts/validate.sh
   

Declarative Configuration

Prefer the repo-local YAML/JSON workflow for full ES coverage:

bash skills/splunk-enterprise-security-config/scripts/setup.sh \
  --spec skills/splunk-enterprise-security-config/templates/es-config.example.yaml \
  --mode preview

Supported modes:

• preview: render the normalized spec, actions, diagnostics, and handoff items.
• apply: perform only supported writes. This mode requires --apply.
• validate: compare planned objects against live Splunk state and execute declarative validation.searches through read-only search export.
• inventory: collect current ES app, index, content, and integration state.
• export: emit a starter brownfield spec from live inventory.

The spec supports these top-level sections:

• baseline - lookup-order, managed roles, optional all-index creation
• indexes - core/UEBA/PCI/exposure/Mission Control/DLX/custom indexes
• roles - ES roles with allowed/default indexes and capabilities
• data_models - ES acceleration + CIM constraint macros
• assets, identities - lookup definitions, optional guarded local lookup uploads, ACL metadata, explicit builder-style scheduled searches, and identity-manager inputs
• threat_intel - native threatlist:// feeds, guarded CSV/STIX/OpenIOC uploads, TIM Cloud readiness
• detections - existing tuning, custom detections with explicit SPL, optional correlation_metadata writes to correlationsearches.conf, drilldown fields, plus optional acl / permissions that POST to /saved/searches/<name>/acl with sharing, owner, and perms
• risk - risk factors and risk-rule saved searches
• urgency - urgency.conf severity-priority matrix
• adaptive_response - alert_actions.conf defaults for notable, risk, finding, rba, sendalert, email, webhook
• notable_suppressions - notable_suppressions.conf stanzas
• log_review - legacy spec keys for statuses / dispositions / settings. ES 8.x routes both statuses and dispositions to reviewstatuses.conf under SA-ThreatIntelligence, and log_review.conf [incident_review] settings to the same app
• review_statuses - modern ES 8.x entry point for notable and investigation review-status stanzas in reviewstatuses.conf
• use_cases - ES Use Case library / navigator entries
• governance - regulatory frameworks such as PCI, NIST, CIS
• macros - arbitrary macros.conf entries (beyond CIM constraints)
• eventtypes - savedsearches.conf eventtypes with tag lists
• tags - Splunk fvtags for field-value tagging
• navigation - app navigation XML (data/ui/nav)
• glass_tables - Glass Tables / ES dashboards as data/ui/views; XML that looks like it contains secret material auto-downgrades to a handoff action
• workflow_actions - analyst right-click actions (data/ui/workflow-actions)
• kv_collections - custom KV Store collections with field types, accelerated fields, and replication flags
• findings - findings.conf [intermediate_findings] settings
• intelligence_management - TIM Cloud subscription (intelligence_management.conf) plus handoff for tenant pairing
• es_ai_settings - ES 8.5 AI assistant and triage settings under Mission Control
• dlx_settings - dlx-app detection lifecycle tunables
• exposure_analytics - safe toggles for entity-discovery processing searches and A&I population saved searches/macros; source and enrichment-rule objects are create-only guarded when a documented endpoint, conflict_policy: create_only, and explicit apply: true are provided
• content_library - Splunk ES Content Update (DA-ESS-ContentUpdate) install via Splunkbase app 3449 on Enterprise apply, optional app-id overrides, ESCU subscription, story/detection enablement, content-pack toggles
• mission_control, integrations, ta_for_indexers, content_governance, package_conf_coverage, validation

Writes require --apply; preview, inventory, export, and validate are read-only. Custom detections require explicit SPL and default to disabled unless enabled: true is present. Mission Control API writes require apply: true on the individual object. Secret-dependent integrations produce handoff steps and must use file-path-only inputs such as secret_file, password_file, token_file, or cert_bundle_file rather than inline secrets. Integration preflight: true validates non-secret fields, app readiness, Splunk REST endpoint availability, and secret-file existence/permissions without reading secret contents. OAuth exchange and tenant pairing remain handoffs with required inputs and safe next commands. Glass Table XML that contains literal password/api_key/private-key markers is rejected as a handoff item even without an explicit secret_file reference.

For content_library.install: true, SA-ContentLibrary is treated as bundled/presence-checked unless content_library.app_ids provides an explicit supported app id. Splunk Cloud content-library installation remains a handoff. For lookup and threat-intel uploads, the spec must opt in with apply: true and a local file path; export includes only metadata and ACLs, never file contents. Splunk_TA_ForIndexers deployment is read-only planning by default and overwrite requires explicit deploy intent plus replace_existing: true or an overwrite policy, backup_export, clean conflict checks, and the preview-generated confirm_id. Content rollback/delete and private Mission Control overrides use the same two-phase confirmation pattern. Asset and Identity Builder-style LDAP or cloud-provider sources require an explicit generating_search.search SPL that writes the declared lookup with outputlookup; otherwise the engine emits a handoff so directory profiles, cloud credentials, and generated SPL stay reviewable. Mission Control search.workload_pool maps to mc_search.conf [aq_sid_caching] for analyst queue searches; create the workload pool with the workload management skill before applying that setting. Optional Mission Control RBAC lockdown is guarded behind rbac_lockdown.apply: true. Private Mission Control internals are inventory/export-first. Only the documented safe allowlist, currently mc_search.conf [aq_sid_caching] workload_pool, is applied directly; other private_overrides require private_override: true, apply: true, backup_export, and a matching preview confirm_id. For Splunk Cloud, connection.cloud_support.evidence_package: true creates a Support-ready evidence package for app/index inventory, requested settings, missing prerequisites, and ACS-supported readiness. Unsupported Cloud REST writes and Support-only ES setup are not attempted. Use package_conf_coverage for inventory/export-only coverage of every .conf family shipped in the local ES package, including internal Mission Control, analytic-story, managed-configuration, app-permission, dashboard-mapping, and package-registration files. It does not create write actions.

Scripts

setup.sh

Applies repeatable ES configuration primitives and runs the declarative engine.

Flag	Purpose
--spec PATH	Run the YAML/JSON ES configuration-as-code workflow
--mode preview|apply|validate|inventory|export	Select declarative workflow mode
--apply	Required guard for declarative writes
--output PATH	Write declarative JSON output to a file
--stop-on-error	Explicitly select the default fail-stop policy. The first failed action halts mutation; remaining actions are skipped, and there is no rollback of earlier successes.
--continue-on-error	Explicit high-risk override that keeps mutating after failures. Use only when actions are known to be independent; there is no rollback.
--strict	Fail fast when the spec contains unknown top-level sections (typo guard for valdation:, detentions:, etc.)
--baseline	Convenience shortcut for lookup order, managed roles, all ES indexes, and validation
--all-indexes	Convenience shortcut for core, UEBA, PCI, exposure, Mission Control, and DLX indexes
--set-lookup-order	Set limits.conf [lookup] enforce_auto_lookup_order=true
--set-managed-roles ROLES	Set App Permissions Manager managed roles
--create-core-indexes	Create core ES indexes through the platform index helper
--create-ueba-indexes	Create UEBA indexes through the platform index helper
--create-pci-indexes	Create PCI indexes through the platform index helper
--create-exposure-indexes	Create exposure analytics indexes
--create-mission-control-indexes	Create Mission Control indexes bundled with ES 8.x
--create-dlx-indexes	Create detection lifecycle/confidence indexes
--max-size-mb N	maxTotalDataSizeMB for newly created indexes
--enable-dm NAME	Enable an ES dm_accel_settings://NAME stanza
--disable-dm NAME	Disable acceleration for an ES data model stanza
--validate	Run validation after changes

No flags means validate only.

Combining declarative and imperative phases

When --spec, --mode, --apply, or --output is set together with imperative shortcuts (--baseline, --create-core-indexes, --set-lookup-order, etc.), the script runs the declarative phase first, then runs the imperative phase in the same invocation. Earlier versions silently dropped the imperative phase. Validation, when requested, runs once at the end of both phases.

Validation searches

validation.searches entries accept expect_rows (default true) and min_event_count (default 1). A search that returns fewer events than the minimum is reported as ok: false with a reason. Set expect_rows: false for "did the saved search even run?" smoke tests that should pass on zero results.

Correlation metadata app discovery

detections[].correlation_metadata.app is now optional. When omitted, the engine probes a documented list of SA-* and DA-ESS-* apps at apply time and POSTs to the first app where the correlation rule resolves. Override the search order via correlation_metadata.app_search_order.

validate.sh

Read-only checks for:

• ES app availability and KV Store
• Core, UEBA, PCI, exposure analytics, Mission Control, and detection lifecycle index presence
• enforce_auto_lookup_order
• App Permissions Manager managed roles
• Data model acceleration enforcement stanzas
• Asset/identity and threat intelligence KV Store collections
• Detection, notable, risk, and threat activity smoke searches

Additional Resources

• reference.md — configuration coverage map, official source links, index checklist, and validation searches.

MCP Tools

This skill includes checked-in, read-only Splunk MCP custom tools generated from mcp_tools.source.yaml.

Validate or regenerate the tool artifact:

python3 skills/shared/scripts/mcp_tools.py validate skills/splunk-enterprise-security-config
python3 skills/shared/scripts/mcp_tools.py generate skills/splunk-enterprise-security-config

Load the tools into Splunk MCP Server:

bash skills/splunk-enterprise-security-config/scripts/load_mcp_tools.sh

The loader uses the supported /mcp_tools REST batch endpoint by default. Use --allow-legacy-kv only for older MCP Server app versions that lack that endpoint.