docs: Add Scalekit integration for authenticated tool calling

[saif-at-scalekit (saif-at-scalekit)]

Summary

Add Scalekit as a LangSmith integration under Developer tools.

Scalekit provides authenticated tool calling for AI agents — 60+ pre-built connectors
(Gmail, Slack, Salesforce, GitHub, Notion, etc.) with delegated OAuth, token management, and native LangChain
StructuredTool support. When LangSmith tracing is enabled, Scalekit tool calls appear as tool spans automatically.

Files changed

• src/langsmith/trace-with-scalekit.mdx — new integration page (SDK adapter + MCP methods)
• src/images/providers/light/scalekit.svg — provider icon (light mode)
• src/images/providers/dark/scalekit.svg — provider icon (dark mode)
• src/langsmith/integrations.mdx — added Scalekit card to Developer tools grid
• src/docs.json — added nav entry

Testing

Code examples verified with traces received in LangSmith UI.
See scalekit-developers/agent-auth-examples#3 for the tested script.

References

• Scalekit AgentKit docs
• Scalekit LangChain guide
• Sample LangChain agent repo
• Package: scalekit-sdk-python

[github-actions]

Thanks for opening a docs PR, saif-at-scalekit (@saif-at-scalekit)! When it's ready for review, please add the relevant reviewers:

• @katmayb or @fjmorris (LangSmith)

[github-actions]

❌ Import check failed

This PR contains imports from langchain_core that should be imported from langchain instead.

Detailed issues

Analyzing diff for import issues...
 Found 2 import issues:

File: src/langsmith/trace-with-scalekit.mdx
Line: 97
Issue: Import from langchain.messages instead of langchain_core.messages
Current:   from langchain_core.messages import HumanMessage, ToolMessage
Suggested: from langchain.messages import HumanMessage, ToolMessage
--------------------------------------------------------------------------------
File: src/langsmith/trace-with-scalekit.mdx
Line: 147
Issue: Import from langchain.messages instead of langchain_core.messages
Current:   from langchain_core.messages import HumanMessage, ToolMessage
Suggested: from langchain.messages import HumanMessage, ToolMessage
--------------------------------------------------------------------------------

 Found 2 import issues that need to be fixed.

Why this is a problem

The langchain package re-exports many modules and classes from langchain_core. When possible, imports should use langchain instead of langchain_core for:

• Better user experience (single import source)
• Consistency across documentation
• Reduced cognitive load for users

How to fix

Replace the imports as suggested above. For example:

• ❌ from langchain_core.messages import HumanMessage
• ✅ from langchain.messages import HumanMessage

🤖 Automated check

This check is based on the latest analysis of langchain re-exports from langchain_core.

[saif-at-scalekit (saif-at-scalekit)]

Addressed the import issue that's pointed out by github actions.

[saif-at-scalekit (saif-at-scalekit)]

Kathryn May (@katmayb) / Florence Morris (@fjmorris) – I’ve gone through other PRs and the contribution guidelines to this pull request. Since this is my first time working on this PR and on #3963, I may have missed a few things. Please let me know if there’s anything I can adjust to make this PR as clean as possible and straightforward for you to review and merge.

[Kathryn May (katmayb)]

Kathryn May (Kathryn May (@katmayb)) / Florence Morris (Florence Morris (@fjmorris)) – I’ve gone through other PRs and the contribution guidelines to this pull request. Since this is my first time working on this PR and on #3963, I may have missed a few things. Please let me know if there’s anything I can adjust to make this PR as clean as possible and straightforward for you to review and merge.

Thanks saif-at-scalekit (@saif-at-scalekit), I'll take a look at this today!

[Kathryn May (katmayb)]

Hi saif-at-scalekit (@saif-at-scalekit) , Thanks so much for putting this together!

After reviewing, we've decided this doesn't quite fit our LangSmith integrations content lineup. Because Scalekit returns standard LangChain StructuredTool objects, the tracing here is LangChain's default behavior; any tool bound to a LangChain model traces automatically when LANGCHAIN_TRACING_V2=true.

This is useful content for using Scalekit with LangChain, which may work really well in the Scalekit docs.

Thanks again, and let me know if you have any questions!

[saif-at-scalekit (saif-at-scalekit)]

I see Kathryn May (@katmayb). There’s also room for the Scalekit SDK to surface a dedicated interface, either something like ScalekitCallbackHandler(BaseCallbackHandler) or langsmith.RunTree / @traceable to emit custom spans, attach richer metadata, or visualize the internal Scalekit tool-calling pipeline. I’ve captured this as product feedback, and for now this guidance will ship in the Scalekit docs.

I assume #3963 is still helpful?

I will close this pull request. Really appreciate feedback.