Skip to main content
GET
Get eval

Authorizations

Authorization
string
header
required

Use a JWT bearer token or a Veryfront API key in the Authorization header.

Path Parameters

project_reference
string
required

Project slug, UUID, or domain that identifies the project.

eval_id
string
required
Minimum string length: 1

Query Parameters

runtime_target_kind
enum<string> | null
Available options:
main_branch,
environment,
preview_branch,
null
target_environment_id
string<uuid> | null
target_branch_id
string<uuid> | null
source_target_kind
enum<string>
Available options:
project,
main_branch,
environment,
preview_branch

Response

The eval definition from the selected source target.

Eval definition.

id
string
required

Eval identifier.

name
string
required

Human-readable eval name.

description
string | null
required

Short description of what the eval checks, or null.

target_kind
enum<string>
required

What the eval runs against: always agent.

Available options:
agent
target
string
required

Identifier of the agent the eval runs against.

source_path
string
required

Source file path where the eval definition lives.

export_name
string
required

Name of the export that declares the eval in its source file.

dataset
object
required

The examples the eval runs.

metrics
object[]
required

The metrics that score each example.

repetitions
integer
required

How often each example runs.

tags
string[]
required

Labels of the eval.

metadata
object
required

Free-form JSON object stored with the eval.

editable_fields
enum<string>[]
required

Fields that update_eval can change.

Available options:
name,
description,
target,
dataset,
repetitions,
tags,
metadata,
metrics,
check
dynamic_fields
enum<string>[]
required

Fields computed in source, which update_eval cannot change.

Available options:
name,
description,
target,
dataset,
repetitions,
tags,
metadata,
metrics,
check
capabilities
string[]
required

Eval capabilities the caller holds on the project.