Splunk core -> python3.13

Hi all,

Thank you for your efforts in getting some work done in the SDK.

Unfortunately, the requirement for Python 3.13, driven by the libraries used in the ai subpackage, renders all new SDK releases unusable for app developers. This is simply because there are multiple still-supported Splunk Enterprise versions that cannot use Python 3.13.

https://www.splunk.com/en_us/legal/splunk-software-support-policy.html#core

I propose creating a second project for the AI-related work and keeping splunk-sdk-python focused on its core purpose. Then you could do something like:

# splunk-sdk
[project]
name = "splunk-sdk"
requires-python = ">=3.9"

[project.optional-dependencies]
ai = [
  "splunk-sdk-ai==1.0.0",
]

# splunk-sdk-ai
[project]
name = "splunk-sdk-ai"
requires-python = ">=3.13"

dependencies = [
  "splunk-sdk>=3.0.0",
  "httpx==0.28.1",
  "langchain>=1.2.18",
  "mcp>=1.27.1",
  "pydantic>=2.13.4",
]

With this, it would be possible to just install the required package for 3.9 as well as pip install "splunk-sdk[ai]" that would fail on python3.9.

Best,
Simon

Start by reviewing the current dependency metadata and the ai subpackage to understand how Python 3.13 is imposed on the SDK. The proposed outcome is a core package that remains usable on Python 3.9 and a separately installable AI package with the newer requirement; the issue does not name specific files or tests.