{
  "registry_version": "1.0.0",
  "data": {
    "title": "Technical Screen / Live Assessment",
    "status": "active",
    "evidence_ids": [
      "evidence.employment_interview_reliability_new_meta_analytic_estimates_by_structure_and_format"
    ],
    "content": "# Technical Screen / Live Assessment\n\nSynchronous technical evaluation assessing core technical competencies, problem-solving speed, and system architecture fundamentals.\n\n### Pass Condition\nThe candidate's total on the assessing engineer's rubric is at or above the threshold to advance.",
    "aliases": [
      "B-005"
    ],
    "id": "bar.technical_screen_live_assessment",
    "type": "barrier",
    "stage": "technical",
    "order": 7,
    "precedes": [
      "bar.take_home_work_sample_evaluation",
      "bar.hiring_manager_in_depth_review"
    ],
    "description": "Synchronous technical evaluation assessing core technical competencies, problem-solving speed, and system architecture fundamentals.",
    "pass_condition": "The candidate's total on the assessing engineer's rubric is at or above the threshold to advance.",
    "evidence_level": "proven",
    "specimens": [
      {
        "kind": "note",
        "label": "The rubric the engineer scored against",
        "lines": [
          {
            "text": "Correctness: 4/5 · Complexity reasoning: 3/5 · Communication: 4/5 · Testing instinct: 3/5",
            "tell": false
          },
          {
            "text": "Threshold to advance: 14/20. Total: 14.",
            "tell": true
          },
          {
            "text": "Recommendation: advance, weak yes.",
            "tell": false
          }
        ],
        "reading": "The gate is a sum against a line. A weak yes and a strong no differ by two points in one column, and the candidate never sees the columns."
      }
    ],
    "perspectives": [
      {
        "actor": "actor.candidate",
        "sees": "An exercise, a conversation about it, and an outcome. Not the columns, not their weights, not the line the total is compared with.",
        "reads": "The hour resolves into a single number that is never shown, so a miss by one point and a clear fail arrive in the same wording.",
        "does": "Reconstructs the session from memory and guesses which part of it cost the points, with nothing in the outcome to check the guess against."
      },
      {
        "actor": "actor.hiring_manager",
        "sees": "Four column scores, a total, and the assessing engineer's recommendation. Not what the same candidate showed in rounds they did not attend.",
        "reads": "A total on the line is an advance with doubt attached. On the company's own account, a bad hire is visible for a year and carries a name while a missed hire carries none — which is why the threshold holds. On a client account the asymmetry can invert: an unfilled billable seat loses revenue every week, and the same market that runs this five-round loop also makes same-week offers.",
        "does": "Holds the threshold where it is and passes the doubt into the next round rather than settling it here."
      },
      {
        "actor": "actor.recruiter",
        "sees": "The recommendation and, at most, the total. The columns and what was said in the room are not part of what is handed over.",
        "reads": "The outcome is settled, and the reason for it stays with the panel.",
        "does": "Writes the message from the recommendation, which is the level of detail available to write from."
      }
    ]
  }
}