Splunk Enterprise SDK for Python

The Splunk Enterprise Software Development Kit (SDK) for Python is intended to be the primary way for developers to communicate with the Splunk platform's REST API.
You may be asking:
Getting started
Requirements
Installing the SDK
Using uv:
uv init
uv add splunk-sdk
uv sync
If you prefer not using uv, using pip should work as expected:
python3 -m venv .venv
source .venv/bin/activate
python3 -m pip install splunk-sdk
Including external dependencies in your App
Because of the way Splunk Apps are built, you need to install your external dependencies to bin/lib/ for the App to work. Then in your App script files:
sys.path.insert(0, os.path.join(os.path.dirname(__file__), "lib"))
Packaging your App
An example workflow to package your App for uploading to Splunk (expect to modify it heavily):
python3 -m pip install . \
--target bin/lib/ \
# Needs to match the platform Splunk is built and
# ran on, NOT the one you're writing your App on
--platform manylinux2014_aarch64 \
--only-binary=:all:
gtar --transform='s,^,<your_app>/,' \
--exclude="__pycache__" \
-czf dist/<your_app>.tgz \
bin default
# `<your_app>.tgz` should be now ready in `dist/`!
See docs for more details.
Using SDK in Splunk Apps
The easiest and most effective way of learning how to use this library should be reading through the apps in our test suite, as well as the splunk-app-examples repository. They show how to programmatically interact with the Splunk platform in a variety of scenarios - from basic metadata retrieval, one-shot searching and managing saved searches to building complete applications with modular inputs and custom search commands.
For details, see the examples using the Splunk Enterprise SDK for Python on the Splunk Developer Portal, as well as the Splunk Enterprise SDK for Python Reference
Using AI in Splunk Apps
You can now leverage AI capabilities within your Splunk Apps using the splunklib.ai package. Take a look at its README to find out how to enhance your Apps agentic behaviour, custom tools and more.
Connecting to a Splunk Enterprise instance
Using a username/password combo
import splunklib.client as client
service = client.connect(host=<HOST_URL>, username=<USERNAME>, password=<PASSWORD>, autologin=True)
Using a bearer token
import splunklib.client as client
service = client.connect(host=<HOST_URL>, splunkToken=<BEARER_TOKEN>, autologin=True)
Using a session key
import splunklib.client as client
service = client.connect(host=<HOST_URL>, token=<SESSION_KEY>, autologin=True)
Creating Custom Search Commands
TODO: Link docs about this
Accessing instance metadata in CSCs
- The
service metadata object is created from the splunkd URI and session key passed to the command invocation the search results info file and is available in MyCommand.generate/transform/stream/reduce methods depending on the Custom Search Command.
from splunklib.searchcommands import StreamingCommand
class MyCommand(StreamingCommand):
def get_metadata(self):
# Access instance metadata
service = self.service
# Access Splunk service info
info = service.info
# [...]
Field retention in CSC
When working with custom search commands such as Custom Streaming Commands or Custom Generating Commands, we may need to add new fields to the records based on certain conditions. Structural changes like this may not be preserved.
If you're having issues with field retention, make sure to use add_field(record, fieldname, value) method from SearchCommand to add a new field and value to the record.
class CustomStreamingCommand(StreamingCommand):
def stream(self, records):
for index, record in enumerate(records):
if index % 1 == 0:
- record["odd_record"] = "true"
+ self.add_field(record, "odd_record", "true")
yield record
Using a helper method to generate events properly
- Generating Custom Search Command is used to generate events using SDK code.
- Make sure to use
gen_record() method from SearchCommand to add a new record and pass event data as comma-separated key=value pairs (mentioned in below example).
@Configuration()
class GeneratorTest(GeneratingCommand):
def generate(self):
- yield {'_time': time.time(), 'one': 1}
- yield {'_time': time.time(), 'two': 2}
+ yield self.gen_record(_time=time.time(), one=1)
+ yield self.gen_record(_time=time.time(), two=2)
Modular Inputs Example App
Go here to find out more about setting up a Modular Input.
Accessing instance metadata in scripts
- The
service metadata object is created from the splunkd URI and session key passed to the command invocation on the modular input stream respectively, and is available as soon as the <script_name>.stream_events() or <script_name>.validate_input() method is called.
from splunklib.modularinput import Script
class MyScript(Script):
def stream_events(self, inputs, ew):
# Access instance metadata
service = self.service
# Access Splunk service info
info = service.info
# [...]
Accessing Modular Inputs' metadata
Contributions
We welcome all contributions!
If you would like to contribute to the SDK, see Contributing to Splunk. For additional guidelines, see CONTRIBUTING.
Setting up a development environment
Make sure you have uv and docker/podman installed and available in your $PATH. Run make install to get a virtualenv set up or updated. After activating it with source .venv/bin/activate you should be ready to go!
Creating an .env file
To connect to Splunk Enterprise, many of the SDK examples and unit tests take command-line arguments that specify values for the host, port, and authentication. For convenience during development, you can store these arguments as key-value pairs in an .env file.
A file called .env.template exists in the root of this repository. Duplicate it as .env, then adjust it to your match your environment.
WARNING: The .env file isn't part of the Splunk platform. This is not the place for production credentials!
Testing
This repository contains a suite of unit and integration tests.
# Run entire test suite:
make test
# Run only the unit tests:
make test-unit
Integration tests
The integration suite requires docker/podman to work.
NOTE: Before running the integration tests, make sure the instance of Splunk you are testing against doesn't have new events being dumped continuously into it. Several of the tests rely on a stable event count. It's best to test against a clean install of Splunk but if you can't, you should at least disable the *NIX and Windows apps.
# This command starts a Splunk Docker container
# and waits until it reaches an operational state.
SPLUNK_VERSION=latest make docker-start
# Run the integration tests:
make test-integration
Do not run the test suite against a production instance of Splunk! It will run just fine with the free Splunk license.
Enabling logging in splunklib
The default level is WARNING, which means that only events of this level and above will be visible
To change a logging level we can call setup_logging() method and pass the logging level as an argument.
Optionally, you can also provide a custom log and date format string. When in doubt, always refer to the source code.
import logging
from splunklib import setup_logging
# To see debug and above level logs
setup_logging(logging.DEBUG)
Documentation and resources
Community
Stay connected with other developers building on the Splunk platform.
Support
-
You will be granted support if you or your company are already covered under an existing maintenance/support agreement. Submit a new case in the Support Portal and include at least Splunk Enterprise SDK for Python in the subject line.
If you are not covered under an existing maintenance/support agreement, you can find help through the broader community at Splunk Answers.
-
Splunk will NOT provide support for SDKs if the core library (code in the /splunklib directory) has been modified. If you modify the SDK and want support, try finding it with the broader community, e.g. Splunk Answers.
That said, we'd also like to know why you felt the need to modify the core library, so please send feedback and file any issues in our GitHub Issues.