Kibana Fleet APIs
You can find details for all available Fleet API endpoints in our generated Kibana API docs.
In this section, we provide examples of some commonly used Fleet APIs.
You can run Fleet API requests through the Kibana Console.
Open the Kibana menu and go to Management → Dev Tools.
In your request, prepend your Fleet API endpoint with
kbn:, for example:GET kbn:/api/fleet/agent_policies
For more detail about using the Kibana Console refer to Run API requests.
Authentication is required to send Fleet API requests. For more information, refer to Authentication.
To create a new agent policy in Fleet, call POST /api/fleet/agent_policies.
This cURL example creates an agent policy called Agent policy 1 in the default namespace.
curl --request POST \
--url 'https://my-kibana-host:9243/api/fleet/agent_policies?sys_monitoring=true' \
--header 'Accept: */*' \
--header 'Authorization: ApiKey yourbase64encodedkey' \
--header 'Cache-Control: no-cache' \
--header 'Connection: keep-alive' \
--header 'Content-Type: application/json' \
--header 'kbn-xsrf: xxx' \
--data '{
"name": "Agent policy 1",
"description": "",
"namespace": "default",
"monitoring_enabled": [
"logs",
"metrics"
]
}'
To save time, you can use Kibana to generate the API request, then run it from the Dev Tools console.
- Go to Fleet → Agent policies.
- Click Create agent policy and give the policy a name.
- Click Preview API request.
- Click Open in Console and run the request.
Example response:
{
"item": {
"id": "2b820230-4b54-11ed-b107-4bfe66d759e4",
"name": "Agent policy 1",
"description": "",
"namespace": "default",
"monitoring_enabled": [
"logs",
"metrics"
],
"status": "active",
"is_managed": false,
"revision": 1,
"updated_at": "2022-10-14T00:07:19.763Z",
"updated_by": "1282607447",
"schema_version": "1.0.0"
}
}
- Make a note of the policy ID. You’ll need the policy ID to add integration policies.
To create an integration policy (also known as a package policy) and add it to an existing agent policy, call POST /api/fleet/package_policies.
You can use the Fleet API to Create and customize an Elastic Defend policy.
This cURL example creates an integration policy for Nginx and adds it to the agent policy created in the previous example:
curl --request POST \
--url 'https://my-kibana-host:9243/api/fleet/package_policies' \
--header 'Authorization: ApiKey yourbase64encodedkey' \
--header 'Content-Type: application/json' \
--header 'kbn-xsrf: xx' \
--data '{
"name": "nginx-demo-123",
"policy_id": "2b820230-4b54-11ed-b107-4bfe66d759e4",
"package": {
"name": "nginx",
"version": "1.5.0"
},
"inputs": {
"nginx-logfile": {
"streams": {
"nginx.access": {
"vars": {
"tags": [
"test"
]
}
},
"nginx.error": {
"vars": {
"tags": [
"test"
]
}
}
}
}
}
}'
To save time, you can use Kibana to generate the API call, then run it from the Dev Tools console.
Go to Integrations, select an Elastic Agent integration, and click Add <Integration>.
Configure the integration settings and select which agent policy to use.
Click Preview API request.
If you’re creating the integration policy for a new agent policy, the preview shows two requests: one to create the agent policy, and another to create the integration policy.
Click Open in Console and run the request (or requests).
To find out which inputs, streams, and variables are available for an integration, go to Integrations, select an Elastic Agent integration, and click API reference.
Example response (truncated for readability):
{
"item" : {
"created_at" : "2022-10-15T00:41:28.594Z",
"created_by" : "1282607447",
"enabled" : true,
"id" : "92f33e57-3165-4dcd-a1d5-f01c8ffdcbcd",
"inputs" : [
{
"enabled" : true,
"policy_template" : "nginx",
"streams" : [
{
"compiled_stream" : {
"exclude_files" : [
".gz$"
],
"ignore_older" : "72h",
"paths" : [
"/var/log/nginx/access.log*"
],
"processors" : [
{
"add_locale" : null
}
],
"tags" : [
"test"
]
},
"data_stream" : {
"dataset" : "nginx.access",
"type" : "logs"
},
"enabled" : true,
"id" : "logfile-nginx.access-92f33e57-3165-4dcd-a1d5-f01c8ffdcbcd",
"release" : "ga",
"vars" : {
"ignore_older" : {
"type" : "text",
"value" : "72h"
},
"paths" : {
"type" : "text",
"value" : [
"/var/log/nginx/access.log*"
]
},
"preserve_original_event" : {
"type" : "bool",
"value" : false
},
"processors" : {
"type" : "yaml"
},
"tags" : {
"type" : "text",
"value" : [
"test"
]
}
}
},
{
"compiled_stream" : {
"exclude_files" : [
".gz$"
],
"ignore_older" : "72h",
"multiline" : {
"match" : "after",
"negate" : true,
"pattern" : "^\\d{4}\\/\\d{2}\\/\\d{2} "
},
"paths" : [
"/var/log/nginx/error.log*"
],
"processors" : [
{
"add_locale" : null
}
],
"tags" : [
"test"
]
},
"data_stream" : {
"dataset" : "nginx.error",
"type" : "logs"
},
"enabled" : true,
"id" : "logfile-nginx.error-92f33e57-3165-4dcd-a1d5-f01c8ffdcbcd",
"release" : "ga",
"vars" : {
"ignore_older" : {
"type" : "text",
"value" : "72h"
},
"paths" : {
"type" : "text",
"value" : [
"/var/log/nginx/error.log*"
]
},
"preserve_original_event" : {
"type" : "bool",
"value" : false
},
"processors" : {
"type" : "yaml"
},
"tags" : {
"type" : "text",
"value" : [
"test"
]
}
}
}
],
"type" : "logfile"
},
...
{
"enabled" : true,
"policy_template" : "nginx",
"streams" : [
{
"compiled_stream" : {
"hosts" : [
"http://127.0.0.1:80"
],
"metricsets" : [
"stubstatus"
],
"period" : "10s",
"server_status_path" : "/nginx_status"
},
"data_stream" : {
"dataset" : "nginx.stubstatus",
"type" : "metrics"
},
"enabled" : true,
"id" : "nginx/metrics-nginx.stubstatus-92f33e57-3165-4dcd-a1d5-f01c8ffdcbcd",
"release" : "ga",
"vars" : {
"period" : {
"type" : "text",
"value" : "10s"
},
"server_status_path" : {
"type" : "text",
"value" : "/nginx_status"
}
}
}
],
"type" : "nginx/metrics",
"vars" : {
"hosts" : {
"type" : "text",
"value" : [
"http://127.0.0.1:80"
]
}
}
}
],
"name" : "nginx-demo-123",
"namespace" : "default",
"package" : {
"name" : "nginx",
"title" : "Nginx",
"version" : "1.5.0"
},
"policy_id" : "d625b2e0-4c21-11ed-9426-31f0877749b7",
"revision" : 1,
"updated_at" : "2022-10-15T00:41:28.594Z",
"updated_by" : "1282607447",
"version" : "WzI5OTAsMV0="
}
}
To get a list of valid enrollment tokens from Fleet, call GET /api/fleet/enrollment_api_keys.
This cURL example returns a list of enrollment tokens.
curl --request GET \
--url 'https://my-kibana-host:9243/api/fleet/enrollment_api_keys' \
--header 'Authorization: ApiKey N2VLRDA0TUJIQ05MaGYydUZrN1Y6d2diMUdwSkRTWGFlSm1rSVZlc2JGQQ==' \
--header 'Content-Type: application/json' \
--header 'kbn-xsrf: xx'
Example response (formatted for readability):
{
"items" : [
{
"active" : true,
"api_key" : "QlN2UaA0TUJlMGFGbF8IVkhJaHM6eGJjdGtyejJUUFM0a0dGSwlVSzdpdw==",
"api_key_id" : "BSvR04MBe0aFl_HVHIhs",
"created_at" : "2022-10-14T00:07:21.420Z",
"id" : "39703af4-5945-4232-90ae-3161214512fa",
"name" : "Default (39703af4-5945-4232-90ae-3161214512fa)",
"policy_id" : "2b820230-4b54-11ed-b107-4bfe66d759e4"
},
{
"active" : true,
"api_key" : "Yi1MSTA2TUJIQ05MaGYydV9kZXQ5U2dNWFkyX19sWEdSemFQOUfzSDRLZw==",
"api_key_id" : "b-LI04MBHCNLhf2u_det",
"created_at" : "2022-10-13T23:58:29.266Z",
"id" : "e4768bf2-55a6-433f-a540-51d4ca2d34be",
"name" : "Default (e4768bf2-55a6-433f-a540-51d4ca2d34be)",
"policy_id" : "ee37a8e0-4b52-11ed-b107-4bfe66d759e4"
},
{
"active" : true,
"api_key" : "b3VLbjA0TUJIQ04MaGYydUk1Z3Q6VzhMTTBITFRTmnktRU9IWDaXWnpMUQ==",
"api_key_id" : "luKn04MBHCNLhf2uI5d4",
"created_at" : "2022-10-13T23:21:30.707Z",
"id" : "d18d2918-bb10-44f2-9f98-df5543e21724",
"name" : "Default (d18d2918-bb10-44f2-9f98-df5543e21724)",
"policy_id" : "c3e31e80-4b4d-11ed-b107-4bfe66d759e4"
},
{
"active" : true,
"api_key" : "V3VLRTa0TUJIQ05MaGYydVMx4S06WjU5dsZ3YzVRSmFUc5xjSThImi1ydw==",
"api_key_id" : "WuKE04MBHCNLhf2uS1E-",
"created_at" : "2022-10-13T22:43:27.139Z",
"id" : "aad31121-df89-4f57-af84-7c43f72640ee",
"name" : "Default (aad31121-df89-4f57-af84-7c43f72640ee)",
"policy_id" : "72fcc4d0-4b48-11ed-b107-4bfe66d759e4"
},
],
"page" : 1,
"perPage" : 20,
"total" : 4
}
Use the Get agents API to retrieve a list of currently enrolled Elastic Agents:
curl -X GET 'http://<user>:<pass>@<kibana url>/api/fleet/agents
By default, a maximum of 10,000 agents are returned, with 20 agents listed per page.
The following query returns the same list, showing 10,000 Elastic Agents per page:
curl -X GET 'http://<user>:<pass>@<kibana url>/api/fleet/agents?perPage=10000'
Elastic Stack
Beginning with Elastic Stack version 9.1, the previous query response includes a nextSearchAfter parameter that you can pass in a subsequent call, to retrieve the next page of 10,000 enrolled agents:
curl -X GET 'http://<user>:<pass>@<kibana url>/api/fleet/agents?perPage=10000&searchAfter=<nextSearchAfter>'
Elastic Stack
Beginning with Elastic Stack version 9.1, you can also capture a point-in-time ID (pitId) parameter from the Get agents API response, and use that together with the nextSearchAfter parameter to capture the next page of 10,000 enrolled agents for a specific point in time.
Include the openPit and pitKeepAlive parameters in your initial request:
curl -X GET 'http://<user>:<pass>@<kibana url>/api/fleet/agents?perPage=10000&openPit=true&pitKeepAlive=5m'
You can then use the returned values in a new request to retrieve the next set of 10,000 agents:
curl -X GET 'http://<user>:<pass>@<kibana url>/api/fleet/agents?perPage=10000&searchAfter=<nextSearchAfter>&pitId=<pit id>&pitKeepAlive=5m'