Assessment Providers

How it works

The first step is for Workable and the Assessments Provider to be authenticated. Then, the user can select one from the available/configured tests from the Provider and push a candidate's information over to create a new assessment. When the candidate completes the assessment, the Provider should inform Workable about the results.


There are 2 endpoints that should be provided, with ${BASE_URL} being the root path for both of them:

  1. https://${BASE_URL}/tests which allows the GET method and lists the available tests
  2. https://${BASE_URL}/assessments which allows the POST method and creates an invitation for a candidate

Workable provides an endpoint for receiving the results, see creating an assessment for more info.


1. Workable to Assessments Provider

Every call from Workable towards the Assessments Provider is authenticated using an access token contained in the Authentication HTTP Header. e.g. Authorization: Bearer 78e3be68d98268509ea304a506ed6362d. This token should be provided to the user from the Assessments Provider and configured in the respective Workable account.

2. Assessments Provider to Workable

Regarding the requests originating from the Assessments Provider, an access token will be generated and communicated upon the integration creation. This token should be included in the Authorization Header as well.

The Bearer token is the standard authorization scheme used in OAuth 2

Listing available tests

Workable needs access to the available/configured tests; an authenticated GET is triggered towards the Assessments Provider to fetch them. The endpoint should be formatted as https://${BASE_URL}/tests and the response should be a JSON document with an array of tests. Each test is described through identification and name.


  "tests": [
      "id": 1,
      "name": "Aptitude Test"
      "id": 2,
      "name": "Accounting Test"

One test per job can be selected. See below where the identification becomes handy.

Creating an Assessment

Workable provides users with the ability to create an assessment for a candidate on demand. If the action is called, Workable will POST the Assessments Provider on https://${BASE_URL}/assessments with a payload that looks as follows:

  "test_id": "12345",
  "job_shortcode": "GROOV005",
  "job_title": "Operations Manager",
  "callback_url": "",
  "candidate": {
    "first_name": "Lakita",
    "last_name": "Marrero",
    "phone": "(785)991-6256",
    "email": "[email protected]",
  • test_id: The identification provided in the previous step
  • job_shortcode: (Optional) The job identified in the list jobs step
  • job_title: The job that the candidate applied to. Can be contained in the invitation email.
  • callback_url: The URL in which the Assessment Provider should publish the results (see below)
  • candidate: Candidate info. Can be used in the invitation email.

The Assessment Provider should respond with a 201 status and an identifier for the created assessment.

{ "assessment_id": "2044922" }

Results collection

1. Publishing the Results to Workable

There are 4 statuses for an assessment. The initial status is pending and then the assessment can be completed, rejected or expired. The status depends on the actions of the candidate regarding the assessment. Whenever the status changes, the Assessments Provider should publish the new status to Workable using PUT on the callback_url provided in the assessment creation step.

When the status is completed, the request should include the assessment results.

2. Publishing the Results on an API

Another alternative is for Workable to poll the endpoint https://${BASE_URL}/assessments/:id with the assessment identifier returned in the assessment creation step. This method however is suboptimal since it requires more than one requests until the results are finally fetched. Moreover, a failsafe mechanism should be created in case the invitation was rejected by the candidate.

In both cases the payload provided by the Assessments Provider should be as follows:

  "results_url": "",
  "status": "completed",
  "assessment": {
    "score": "78",
    "grade": "excelled",
    "summary": "This candidate is an excellent prospect.",
    "details": {
        "behavior": {
          "Influence": 97,
          "conscientiousness": 76
        "mental_skills":  {
          "Problem Solving": 82,
          "Aptitude": 91
    "duration": "01:01:17"
  "attachments": [
      "description": "Assessment Report",
      "url": ""

Mandatory fields

  • status: This can take any of the values completed, declined or expired
  • results_url: A link to the fully fledged report on the Provider's site (Mandatory only when status is completed)

Optional fields

  • grade: This is the final conclusion of the assessment, it can take any of the values failed, passed or excelled. This is respectively mapped to no, yes and definitely yes in Workable evaluation system.
  • assessment:
    • score: An overall score, preferably a percentage
    • summary: Small description/evaluation of the assessment
    • details: This can be used to provide a deeper analysis on the candidate's results. It's a JSON with a maximum of two levels nested objects and is presented to the user with minimum formatting, so keys should be descriptive and readable. The structure should follow the key: value format, no arrays should be included and a first level value may be consisted of more objects itself. Any values holding dates should conform to ISO 8601.
    • duration: How long did it take the candidate to complete the assessment, if available
  • attachments: Currently only supporting PDF. A results report of the assessment along with a description, if available. The URL should be publicly accessible.


The endpoints of the Assessments Provider should handle request errors with a status code followed by a readable message. The description message should be short yet verbose enough to facilitate ease of troubleshooting.

TypeStatusExample messages
Authentication401Missing Token
Authentication401Invalid Token
Invalid Request400Invalid JSON
Invalid Request400Invalid field: name should be a string
Invalid Request422Missing field: name should be provided
Conflict409Entity is already updated

A valid response on an Error would be

{ "status": 401, "message": "Invalid Token" }

3. Retrieving a shared link

This is an optional GET endpoint to be implemented by the Assessments Provider. This endpoint serves assessment links, in case when the system supports sharing public links. The URL should look like https://${BASE_URL}/assessments/:id/shared-link and the assessment-id should be the one returned from the assessment creation call.


  "url": "",
  "ttl": "120",
  "ttl_units": "minutes"
  • url: A link to the public shared report on the Provider's site
  • ttl: (Optional) The time this link will be available. If this is not set, the link is assumed to be permanent.
  • ttl_units: The units used for measuring the TTL. Currently supported values: minutes, days and views.