Supercharge your AI with Qualified.io. Manage the entire candidate lifecycle from one prompt.

This Qualified.io MCP Server connects your AI agent right into your technical hiring pipeline. You can manage everything from setting up new coding tests to digging deep into candidate performance metrics—all without leaving your chat window.

You'll find tools that let you control the full assessment lifecycle. Use list_assessments to pull a list of every available assessment type, regardless of whether it’s currently active. Need details on one? Call get_assessment for all the settings related to a specific test.

To get started, you build the test with create_assessment, which initiates an entirely new coding assessment within the system. Once that's set up and you’ve checked the components using list_challenges or get_challenge for individual challenge details, you can make it live by calling publish_assessment. Don't forget to call unpublish_assessment when you need to pull it back offline.

You manage test visibility with archiving. You can move an existing assessment into a paused state using archive_assessment, and if you change your mind, you restore it—but only viewable again through unarchive_assessment. If an assessment needs updating without changing its core status, you've got the update_assessment_result_review tool to modify or refresh any pre-existing review document associated with a result.

Controlling candidate participation is another big chunk of what this server handles. To send out tests, your agent can invite candidates individually using invite_candidates, or you can manage whole groups by calling list_assessment_cohorts and sending bulk invites via invite_candidates_via_cohort. If a candidate drops out before starting, you use cancel_invitation to stop or revoke their specific test invitation.

For managing the larger hiring flow, your agent pulls up all existing assessment groups using list_assessment_cohorts, letting you keep track of everyone in a particular hiring cycle.

Analyzing what candidates actually do is where this server shines. You pull the core data and overall score for any completed test by running get_assessment_result. For a comprehensive deep dive into performance, use get_assessment_result_exhibit; this retrieves detailed, structured exhibit data so you can analyze specific code quality points. To see how multiple people did against each other, your agent pulls up the scores and status for groups of completed tests using list_assessment_results.

You even get a simple overview of all available challenges by calling list_challenges.

Need to adjust a candidate's test status? No sweat. If a candidate needs another shot at it, you use schedule_retry_assessment_result to reset their result for a later attempt. Conversely, if they cheated or the test was invalid, you can immediately invalidate and close out their specific performance record using terminate_assessment_result.

You also have access to get_challenge which retrieves detailed component information for any challenge used in an assessment.

Here's what your agent does when it runs: It combines these tools. For example, you tell it to list the open assessments and then invite a group of people from a specific cohort. Or, you ask it to pull Candidate X’s results, run get_assessment_result for the score, and then use get_assessment_result_exhibit to get the technical proof of how they performed on the code challenges.