.do
AI & Intelligence

evals

Evaluate AI model performance automatically with 100+ built-in metrics. Run quality checks, compare models, detect regressions, and optimize prompts. Ship reliable AI products with confidence.

evals

Evals.do

Overview

The evals primitive is part of the .do platform's semantic API system. It follows the $.Subject.verb.Object pattern for consistent, intuitive interactions across SDK, CLI, API, and MCP interfaces.

Semantic Pattern

All evals operations follow the semantic $.Evals.verb.Object pattern:

import { $, evals } from 'sdk.do'

// Create resources
await $.Evals.create.Resource({ /* config */ })

// Query data
await $.Evals.get.Data({ /* params */ })

// Update state
await $.Evals.update.Status({ /* changes */ })

Quick Start

import { evals } from 'sdk.do'

// Initialize and use the primitive
const result = await evals.operation({
  // configuration parameters
})

Access Methods

The evals primitive is accessible through multiple interfaces:

SDK

TypeScript/JavaScript library for programmatic access:

import { evals } from 'sdk.do'
await evals.method({ /* ... */ })

SDK Documentation

CLI

Command-line interface for terminal operations:

do evals <command> [options]

CLI Documentation

API

REST/RPC endpoints for HTTP integration:

curl -X POST https://api.do/v1/evals \
  -H "Content-Type: application/json" \
  -d '{"key":"value"}'

API Documentation

MCP

Model Context Protocol for AI assistant integration:

Use evals primitives with AI assistants via MCP

MCP Documentation

See other primitives in the ai-intelligence category.