Connect LangSmith to Your AI Agent

Ask the assistant to find a dataset, inspect its examples, or add the test cases you provide.

10 tools API key

Try a LangSmith Task

Copy a prompt into your agent with the oo Skill installed. Add your page links and source notes before running it.

oo Skill LangSmith

Review Dataset Examples

Find a dataset and inspect examples to summarize its inputs, outputs, and available metadata.

/oo LangSmith: Find LangSmith dataset [name], list its examples, read example [ID], and summarize the available inputs and expected outputs.

Save a Test Example

Add your supplied inputs, outputs, metadata, and split to an existing dataset.

/oo LangSmith: Create a LangSmith example in dataset [ID] with inputs [JSON], expected outputs [JSON], metadata [JSON], and split [split].

Start Working with LangSmith

  1. Install OOMOL

    Set up oo CLI and the oo Skill in your agent.

    Install OOMOL →
  2. Connect LangSmith

    Connect a LangSmith personal or service API key, selecting your SaaS region and workspace when needed.

    Connect LangSmith →
  3. Give Your Agent a Task

    Use /oo with the pages to read, the changes to make, and where to save the result.

    Connection Guide →

Choose Your Agent

  • A LangSmith API key with access to the intended workspace.
  • Supply a workspace ID when the key can access multiple workspaces and the task needs one specific workspace.

LangSmith Tools

LangSmith organizes tracing projects and datasets. This connection reads and creates projects and datasets and reads or adds dataset examples; it does not query individual trace runs or execute evaluations.

Workspaces and Projects

Inspect accessible workspaces and tracing projects or create a named project.

Datasets and Examples

Read or create datasets, inspect example records, and add structured test cases.

10 tools

List Workspaces

List LangSmith workspaces visible to the connected API key.

langsmith.list_workspaces

List Projects

List LangSmith tracing projects with optional name and pagination filters.

langsmith.list_projects

Get Project

Get a LangSmith tracing project by ID.

langsmith.get_project

Create Project

Create a LangSmith tracing project.

langsmith.create_project

List Datasets

List LangSmith datasets with optional name, type, and pagination filters.

langsmith.list_datasets

Get Dataset

Get a LangSmith dataset by ID.

langsmith.get_dataset

Create Dataset

Create a LangSmith dataset.

langsmith.create_dataset

List Examples

List LangSmith dataset examples with optional dataset and text filters.

langsmith.list_examples

Get Example

Get a LangSmith dataset example by ID.

langsmith.get_example

Create Example

Create a LangSmith dataset example with JSON inputs, outputs, and metadata.

langsmith.create_example

Frequently Asked Questions

Can I read examples from a dataset version?

Yes. Example reads accept a dataset version date or latest, where applicable.

What do I supply for a new example?

Choose a dataset and provide JSON inputs and outputs appropriate to its data type, with optional metadata and split.

Can I run an evaluation here?

The connection manages dataset and project records. It does not start evaluation runs.

Connect LangSmith

About LangSmith ↗