Skip to main content

Documentation Index

Fetch the complete documentation index at: https://next.gentrace.ai/docs/llms.txt

Use this file to discover all available pages before exploring further.

The pipelines SDK provides programmatic access to Gentrace pipelines. Pipelines are the top-level organizational unit that contains datasets, experiments, and traces.

​
Basic usage

import { init, pipelines } from 'gentrace';

init({
  apiKey: process.env.GENTRACE_API_KEY,
});

// List pipelines
const pipelineList = await pipelines.list();

// Access the pipelines
for (const pipeline of pipelineList.data) {
  console.log(pipeline.slug);
  console.log(pipeline.displayName);
}

​
Overview

The SDK is built by Stainless and provides type-safe access to Gentrace entities. The pipelines object exposes methods to create, retrieve, update, and list pipelines.

​
Pipeline structure

Each pipeline contains:
​
id
string
required
Unique identifier (UUID)
​
slug
string
required
URL-friendly unique identifier
​
displayName
string
Optional human-readable name
​
folderId
string
Optional folder UUID for organization
​
goldenDatasetId
string
UUID of the golden dataset (if set)
​
archivedAt
string
Timestamp when archived (null if active)
​
createdAt
string
required
Creation timestamp
​
updatedAt
string
required
Last update timestamp

​
Resource methods

​
Create a pipeline

const pipeline = await pipelines.create({
  slug: 'customer-support-ai',
  displayName: 'Customer Support AI Pipeline',
  folderId: 'folder-uuid', // optional
});

​
Retrieve a pipeline

const pipeline = await pipelines.retrieve('pipeline-id');
console.log(pipeline.slug);

​
Update a pipeline

const pipeline = await pipelines.update('pipeline-id', {
  displayName: 'Updated Pipeline Name',
  isArchived: false,
});

​
List with filters

// Filter by folder or slug
const pipelineList = await pipelines.list({
  folderId: 'folder-id',
  slug: 'customer-support', // exact match or use advanced filters
});

​
Usage in AI workflows

Pipelines are used throughout Gentrace to organize your AI workflows: 
// Use pipeline ID with interaction()
const tracedFunction = interaction('My AI Function', myAIFunction, {
  pipelineId: pipeline.id,
});

// Use pipeline ID with experiment()
experiment(pipeline.id, async () => {
  // Your experiment code
});

​
See also