Skip to main content

create-an-interview-scorecard

Turn role criteria into observable, consistently scored interview evidence with anchored ratings. Use when preparing structured candidate evaluation.

ソース情報

リポジトリ
protosphinx/sphinxstack
ソースの最終更新活動
2026年7月29日 13:55
検出された SKILL.md の言語
英語
スター
3
フォーク
0

インストール方法

デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。

ソースファイルを確認

インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。

ファイルエクスプローラー
2 ファイル

SKILL.md を表示中

SKILL.md
ソースの指示 · 読み取り専用プレビュー
name
create-an-interview-scorecard
category
get-hired
description
Turn role criteria into observable, consistently scored interview evidence with anchored ratings. Use when preparing structured candidate evaluation.
# create-an-interview-scorecard Score evidence against the role, not resemblance to prior hires. Build anchors before seeing candidate performance. ## Procedure 1. Start from approved role outcomes and the hiring process's competency model. 2. Select only competencies that materially distinguish performance and can be assessed in the planned stage. 3. Define observable positive, mixed, insufficient, and concerning evidence for each criterion. 4. Write behaviorally anchored rating levels and an explicit not-observed state. 5. Map every question or work sample to named criteria and remove duplicate or irrelevant assessment. 6. Set evidence-note requirements and prohibit unsupported labels such as culture fit, executive presence, or polish. 7. Define must-have gates, compensating strengths, weighting, and decision rules before interviews. 8. Test the scorecard on representative fictional responses and compare assessor interpretations. 9. Train assessors to score independently before debrief and to record accommodation effects appropriately. ## Failure plan - Do not score accent, appearance, disability, family status, school prestige, similarity, or other irrelevant proxies. - Never convert missing opportunity to demonstrate into a negative score. - If an interview question fails to elicit the intended evidence, fix the assessment rather than guessing. - Keep candidate evidence restricted and retain it only as authorized. ## Done - Every criterion traces to a job outcome and an assessment - Rating anchors, evidence requirements, gates, and not-observed handling are tested - Assessors reach consistent interpretations on a rehearsal case
GitHubで見る