Skip to main content

plan-a-pilot-program

Plan a pilot program with a bounded population, hypotheses, operational controls, success and harm measures, support, stop rules, and scale decision. Use when testing a new service, policy, product, process, or partnership before broader adoption.

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
plan-a-pilot-program
category
start
description
Plan a pilot program with a bounded population, hypotheses, operational controls, success and harm measures, support, stop rules, and scale decision. Use when testing a new service, policy, product, process, or partnership before broader adoption.
# plan-a-pilot-program A pilot should reduce a named uncertainty without exposing participants to ungoverned risk. ## When to use - Use when representative real-world operation is needed before scale. - Do not select only easy participants or call a promotional launch an experiment. ## Procedure 1. State decision, hypotheses, unknowns, alternatives, sponsor, and authority. 2. Define representative population, inclusion, exclusion, recruitment, consent where needed, accessibility, and sample limitations. 3. Bound geography, time, volume, capability, staff, data, integrations, vendors, and financial exposure. 4. Establish baseline, success, quality, safety, fairness, privacy, cost, adoption, and harm measures. 5. Define workflow, training, support, incident, exception, continuity, rollback, and participant communication. 6. Set stop, pause, expand, repeat, modify, and reject criteria before starting. 7. Capture quantitative outcomes, qualitative evidence, operational burden, failures, and missing data. 8. Make the scale decision with limitations, conditions, owner, and next evidence. ## Done - A pilot plan records decision, population, scope, controls, measures, operations, support, stop rules, timeline, and owners - Representativeness, baseline, consent, accessibility, data, safety, support, measure, stop, and scale-decision checks verify learning
GitHubで見る