langchain-ai/docs
1
0
Fork 0
mirror of https://github.com/langchain-ai/docs.git synced 2026-07-22 12:25:55 -04:00
Code Issues 490 Packages Projects Releases Wiki Activity

[GH-ISSUE #477] [langgraph]: No Documentation for Injecting MCP Tools in LangGraph Graph Definitions #34

New Issue
Open
opened 2026-02-17 17:19:02 -05:00 by yindo · 1 comment
yindo commented 2026-02-17 17:19:02 -05:00
Owner
Copy Link
Copy Source

Originally created by @magallardo on GitHub (May 28, 2025).
Original GitHub issue: https://github.com/langchain-ai/docs/issues/477

Issue with current documentation:

I'm trying to define a LangGraph graph where the tools node uses a predefined set of tools sourced from an MCP Server. However, I couldn't find any documentation or examples showing how to inject MCP tools into the graph definition itself.

The current documentation and examples only demonstrate how to:

  • Use MultiServerMCPClient.get_tools() in a client context.
  • Pass the retrieved tools to an agent at runtime.

What I’d like to see is:
A pattern where the LangGraph graph is constructed with MCP tools already available, so that clients invoking the graph don’t need to fetch or provide the tools themselves. Note that the graph will be deployed to Langgraph Server.

Why this is an issue:

  • MultiServerMCPClient.get_tools() is an async function.
  • Graph definition (nodes, connections) is done in a non-async context.
  • This makes it difficult to cleanly inject MCP tools at graph construction time.

While some examples exist, they’re focused on client-side usage, not on building self-contained graphs with pre-configured tools.

Ask:
Can you provide guidance or add documentation for:

  • How to load MCP tools during graph setup (possibly using an async bootstrap step).
  • A recommended pattern for embedding MCP-sourced tools in LangGraph graph definitions.

Any help or best practices on resolving this would be greatly appreciated!

Thanks!

Idea or request for content:

No response

Originally created by @magallardo on GitHub (May 28, 2025). Original GitHub issue: https://github.com/langchain-ai/docs/issues/477 ### Issue with current documentation: I'm trying to define a LangGraph graph where the tools node uses a predefined set of tools sourced from an MCP Server. However, I couldn't find any documentation or examples showing how to inject MCP tools into the graph definition itself. The current documentation and examples only demonstrate how to: - Use MultiServerMCPClient.get_tools() in a client context. - Pass the retrieved tools to an agent at runtime. What I’d like to see is: A pattern where the LangGraph graph is constructed with MCP tools already available, so that clients invoking the graph don’t need to fetch or provide the tools themselves. Note that the graph will be deployed to Langgraph Server. Why this is an issue: - MultiServerMCPClient.get_tools() is an async function. - Graph definition (nodes, connections) is done in a non-async context. - This makes it difficult to cleanly inject MCP tools at graph construction time. While some examples exist, they’re focused on client-side usage, not on building self-contained graphs with pre-configured tools. Ask: Can you provide guidance or add documentation for: - How to load MCP tools during graph setup (possibly using an async bootstrap step). - A recommended pattern for embedding MCP-sourced tools in LangGraph graph definitions. Any help or best practices on resolving this would be greatly appreciated! Thanks! ### Idea or request for content: _No response_
yindo added the langgraphexternal labels 2026-02-17 17:19:02 -05:00
yindo commented 2026-02-17 17:19:02 -05:00
Author
Owner
Copy Link
Copy Source

@sydney-runkle commented on GitHub (Jun 11, 2025):

Definitely something we want to have more support for, thanks for bringing this to our attention. We're working on a docs restructure now!

@sydney-runkle commented on GitHub (Jun 11, 2025): Definitely something we want to have more support for, thanks for bringing this to our attention. We're working on a docs restructure now!
yindo changed title from [langgraph]: No Documentation for Injecting MCP Tools in LangGraph Graph Definitions to [GH-ISSUE #477] [langgraph]: No Documentation for Injecting MCP Tools in LangGraph Graph Definitions 2026-06-05 17:24:40 -04:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
bug
deepagents
docs-infra
external
external
help wanted
integration
internal
internal
javascript
js/ts
langchain
langgraph
langsmith
needs-title
pull-request
Mirrored from GitHub Pull Request
 python
reference
No Label external langgraph
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langchain-ai/docs#34
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?