kibana-streams

Kibana Streams

Read stream metadata, settings, and attachments, and manage stream lifecycle (enable, disable, resync) via the Kibana Streams REST API. Streams are an experimental way to manage data in Kibana — expect API and behavior changes. This skill covers read operations and lifecycle only; create, update, delete, fork, and other mutating operations may be added in a later version.

For detailed endpoints and parameters, see references/streams-api-reference.md.

When to use

• Listing all streams or getting a single stream's definition and metadata
• Reading a stream's ingest settings, or the ES|QL definition of a query stream
• Listing attachments (dashboards, rules, SLOs) linked to a stream
• Enabling, disabling, or resyncing streams

Prerequisites

Use the space-scoped path /s/{space_id}/api/streams when operating in a non-default space. For role configuration (Kibana feature privileges and Elasticsearch-level permissions), refer to Streams required permissions.

API base and headers

• Base path: GET or POST to <kibana_url>/api/streams (or /s/<space_id>/api/streams for a space).
• Read operations: Typically do not require extra headers; follow the official API docs for each endpoint.
• Lifecycle operations: POST /api/streams/_disable, _enable, and _resync are mutating — send kbn-xsrf: true (or equivalent) as required by your Kibana version.

Operations (read + lifecycle)

Read

Lifecycle

Path parameters: {name} and {streamName} are the stream identifier (same value; the API docs use both names).

Lifecycle and retention (ingest settings)

Ingest settings (GET /api/streams/{name}/_ingest) expose two separate lifecycle areas:

• Stream lifecycle (ingest.lifecycle) — Controls how long the stream's data is retained. Use lifecycle.dsl.data_retention (e.g. "30d") for explicit retention, or lifecycle.inherit for child streams. This is what users usually mean when they ask to "set retention", "update retention", or "change the stream's retention".
• Failure store lifecycle (ingest.failure_store.lifecycle) — Controls retention of failed documents only (documents that did not process successfully). Users rarely need to change this unless they explicitly mention the failure store or failed-document retention.

When a user asks to set or update retention, target the stream's main lifecycle (lifecycle.dsl.data_retention), not the failure store, unless they specifically ask about failure store or failed documents.

Examples

List streams

curl -X GET "${KIBANA_URL}/api/streams" \
  -H "Authorization: ApiKey <base64-api-key>"

Get a single stream

curl -X GET "${KIBANA_URL

Get query stream settings

Get stream attachments

Disable, enable, or resync streams

Check the Streams API operation pages for request/response bodies (e.g. request body for _disable/_enable/_resync if required).

Guidelines

• When the user asks to set or update retention, assume they mean the stream's data retention (ingest.lifecycle / lifecycle.dsl.data_retention). Do not change only the failure store retention unless they explicitly ask about the failure store or failed documents.
• Other mutating operations (create, update, delete, fork, attachment management, and more) are not supported by this skill. See references/streams-api-reference.md for the full list of deferred operations.
• Significant events and their detection queries are out of scope. Requests about significant events, SigEvents, or Knowledge Indicators belong to the dedicated significant-events skill, which uses the platform.sig_events.* Agent Builder tools. The /api/streams/{name}/queries and /api/streams/{name}/significant_events endpoints are deprecated and pending removal in Kibana — do not reach for them here.
• Do not confuse a query stream with a significant-event query. /api/streams/{name}/_query is the ES|QL definition of a query stream (a stream type, alongside wired and classic); it has nothing to do with detection queries.
• Disabling streams can lead to data loss for wired streams. The disable API deletes wired stream data (classic stream data is preserved). Before calling disable, warn the user and confirm they understand the risk (and have backed up or no longer need the data).
• Prefer read operations when the user only needs to inspect stream state; use lifecycle APIs when they need to enable, disable, or resync streams.