Generate Criteria

Create screening criteria from a job description using AI

POST /v1/jobs/{jobId}/criteria/generate

@nova-sdk/api

TypeScript

import { Nova } from "@nova-sdk/api";

const nova = new Nova({
  apiKey: "sk_test_...",
  tenantId: "acme-corp",
});

const result = await nova.jobs.criteria.generate.create({
  jobId: "job_abc123",
  body: {
    jobContext: {
      jobTitle: "Senior Software Engineer",
      companyName: "Acme Corp",
      jobDescription: "We are looking for a senior engineer...",
    },
  },
});

201

{
  "criteria": [
    {
      "id": "<string>",
      "text": "<string>",
      "order": 123,
      "createdAt": "2023-11-07T05:31:56Z"
    }
  ]
}

Creates a new version: If this job already has active criteria, a new criteria version is created and becomes the active version. Historical scores retain the criteria version they were scored against.

When to Use

Generate criteria once per job posting, typically when the job is finalized. This endpoint:

• Analyzes the job description to identify key requirements
• Creates structured screening criteria with importance levels
• Stores the criteria server-side for automatic use during scoring

With or Without Questions

With Clarification Answers

Recommended for best results. Requires a prior call to Generate Clarification Questions.

Clarification answers help calibrate:

• Which skills are MUST_HAVE vs NICE_TO_HAVE
• The target seniority level
• Any context not captured in the job description

Without Answers

Generates criteria directly from the job description. Use this when:

• You want a quick setup without the questions workflow
• The job description is comprehensive and unambiguous
• You’ll refine criteria later using the criteria management endpoints

Answer requirements

If you include answers, also include questionSetId and answer every question in the question set.

Example request

TypeScript SDK

import { Nova } from '@nova-sdk/api';

const nova = new Nova({
  apiKey: process.env.NOVA_API_KEY!,
  tenantId: 'acme-corp',
});

const criteria = await nova.jobs.criteria.generate.create({
  jobId: 'job-123',
  body: {
    jobContext: {
      jobTitle: 'Senior Software Engineer',
      companyName: 'Acme Corp',
      jobDescription: 'We are looking for a Senior Software Engineer with 5+ years experience...',
    },
    questionSetId: 'cmlz26fn5003dwp61u6vmj4mc',
    answers: [
      { id: 'q1', value: 'Senior (5+ years)' },
      { id: 'q2', value: ['Go', 'Kubernetes'] },
    ],
  },
});

Example response

{
  "criteria": [
    {
      "id": "crit_a1b2c3d4e5f6",
      "text": "Strong proficiency in Go for backend development",
      "importance": "MUST_HAVE",
      "order": 0
    }
  ]
}

Latency

10-20 seconds (synchronous)

The AI analyzes the job context to generate contextually relevant criteria. This is a one-time operation per job posting.

Next Step

After generating criteria, you’re ready to Score Applications. Nova loads the job’s active criteria automatically.

Authorizations

Authorization · string · header · required

Use Authorization: Bearer sk_test_* or Authorization: Bearer sk_live_*.

Headers

X-Tenant-Id · string · required

Your customer identifier. Tenants are auto-provisioned on first request.

Idempotency-Key · string

Optional key that makes criteria and library mutation requests safe to retry without duplicating side effects. The namespace is shared across all API keys for the same partner and environment. Reuse the same key when retrying the same request parameters.

Required string length: 1 - 255

Path Parameters

jobId · string · required

Your job identifier (external ID).

Body

application/json

jobContext · object · required

Show child attributes

questionSetId · string

Minimum string length: 1

answers · object[]

Show child attributes

Response

201

application/json

Created

criteria · object[] · required

Show child attributes