Skip to content

Instantly share code, notes, and snippets.

@esz135888

esz135888/acceptance-tests.md

Last active May 24, 2026 03:07
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save esz135888/387580ff96fad9da5d12fdf027f07fe6 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save esz135888/387580ff96fad9da5d12fdf027f07fe6 to your computer and use it in GitHub Desktop.
Download ZIP
PLS AI adoption assistant 30 day command center production pack
Raw
acceptance-tests.md

Acceptance Tests

Production Acceptance

  1. D7 artifact threshold.

    • Given D7 has passed
    • Then each product line has at least one openable artifact URL or the workstream is flagged blocked with owner.

  2. D14 evidence threshold.

    • Given D14 has passed
    • Then at least one real usage or owner response evidence is attached.

  3. D30 project threshold.

    • Given D30 has passed
    • Then total accepted artifacts >= 3, usage evidence >= 3, overdue clearance >= 80%, and a D30 decision exists.

  4. Revenue path is explicit.

    • Given a product line is marked scaled
    • Then it must have revenue hypothesis, pipeline owner and next customer action.

  5. Owner change is allowed.

    • Given a workstream has no artifact by D7 and no evidence by D14
    • Then D30 decision options must include change_owner or pause.

  6. Audit trail exists.

    • Given a D30 decision is recorded
    • Then decided_by, reason, evidence_summary and timestamp are present.

E2E Verification This Round

  • Primary artifact is HTML.
  • Public Gist returns HTTP 200.
  • Gist includes all 13 files.
  • PLS upload-files receives all 13 files.
  • Complete artifacts_json uses stable public refs.
Raw
ai-adoption-assistant-30day-command-center.html
<!doctype html>
<html lang="zh-Hant">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>AI 落地助手 30 天導入作戰台</title>
<style>
:root{--ink:#111827;--muted:#667085;--paper:#f5f7fb;--card:#fff;--line:#d8e0ea;--blue:#2454c6;--green:#08785e;--red:#b13b2c;--amber:#a06000;--violet:#6842bf}
*{box-sizing:border-box}body{margin:0;background:var(--paper);color:var(--ink);font-family:ui-sans-serif,system-ui,-apple-system,BlinkMacSystemFont,"Segoe UI",sans-serif;line-height:1.5}header{background:#fff;border-bottom:1px solid var(--line);padding:30px clamp(18px,4vw,56px)}main{padding:24px clamp(18px,4vw,56px) 48px}
h1{margin:0 0 12px;font-size:clamp(32px,4vw,56px);line-height:1.04;letter-spacing:0;max-width:1160px}h2{margin:0 0 12px;font-size:22px}h3{margin:0 0 6px;font-size:16px}p{margin-top:0}.sub{max-width:1120px;color:var(--muted);font-size:17px}.grid{display:grid;gap:16px}.kpis{grid-template-columns:repeat(4,minmax(0,1fr));margin-top:22px}.two{grid-template-columns:1.05fr .95fr}.three{grid-template-columns:repeat(3,minmax(0,1fr))}.four{grid-template-columns:repeat(4,minmax(0,1fr))}.flow{grid-template-columns:repeat(5,minmax(0,1fr))}
.card{background:var(--card);border:1px solid var(--line);border-radius:8px;padding:18px;box-shadow:0 1px 2px rgba(17,24,39,.04)}.metric{font-size:34px;font-weight:800}.label{color:var(--muted);font-size:13px}.pill{display:inline-flex;border:1px solid var(--line);border-radius:999px;padding:4px 10px;font-size:12px;background:#fff;margin:0 6px 8px 0;white-space:nowrap}.ok{color:var(--green);font-weight:760}.bad{color:var(--red);font-weight:760}.warn{color:var(--amber);font-weight:760}
table{width:100%;border-collapse:collapse;font-size:14px}th,td{text-align:left;padding:10px;border-bottom:1px solid var(--line);vertical-align:top}th{color:var(--muted);font-size:12px;text-transform:uppercase}code{background:#edf3f7;border-radius:4px;padding:1px 5px}.step{border:1px solid var(--line);border-radius:8px;padding:12px;min-height:132px;background:#fbfdff}.step strong{display:block;color:var(--violet);margin-bottom:6px}.source a{color:var(--blue);word-break:break-word}
@media(max-width:960px){.kpis,.two,.three,.four,.flow{grid-template-columns:1fr}h1{font-size:34px}}
</style>
</head>
<body>
<header>
<span class="pill">PLS production delivery pack</span><span class="pill">Solution: system / project / scorecard</span>
<h1>AI 落地助手 30 天導入作戰台</h1>
<p class="sub">把「AI落地助手導入」從逾期訊號變成可驗收 30 天推進系統：整合 AI Coach、AgentERP、FDE 三大產品線，產出可打開成果、3 個以上 action items、真實使用回填證據，最後做 D30 續行/暫停/換 owner/升級系統決策。</p>
<section class="grid kpis">
<div class="card"><div class="metric ok">3</div><div class="label">D30 至少 3 件可打開/可執行成果</div></div>
<div class="card"><div class="metric">1200萬</div><div class="label">季度營收 North Star</div></div>
<div class="card"><div class="metric warn">80%</div><div class="label">逾期待辦清除率目標</div></div>
<div class="card"><div class="metric">D30</div><div class="label">續行 / 暫停 / 換 owner / 升級系統</div></div>
</section>
</header>
<main class="grid">
<section class="grid two">
<div class="card">
<h2>本輪任務</h2>
<p>目前母專案的問題不是缺想法，而是缺「可驗收導入節奏」。本輪把 AI Coach、AgentERP、FDE 拆成同一個 30 天 adoption operating system：每條線都有 owner、D7 artifact、使用證據、營收假設與 D30 決策。</p>
<span class="pill">Owner: Louis</span><span class="pill">Cadence: daily</span><span class="pill">Acceptance: 3 artifacts + 3 evidence</span>
</div>
<div class="card">
<h2>解法選型</h2>
<p><strong>system / project / scorecard / agent</strong>。AI 導入不是一次性文件；需要把產品、客戶、owner、使用證據、營收 pipeline、風險與 D30 決策串成可追蹤系統。</p>
</div>
</section>
<section class="card">
<h2>D1 / D7 / D14 / D30</h2>
<div class="grid four">
<div class="card"><h3>D1</h3><p>建立三線產品矩陣、owner、客戶場景、資料欄位、第一批 action items。</p></div>
<div class="card"><h3>D7</h3><p>每條線至少 1 件可打開成果：AI Coach demo、AgentERP workflow、FDE playbook。</p></div>
<div class="card"><h3>D14</h3><p>取得 1 次真實使用/回填證據，開始測 NPS、願付費率、pipeline 金額。</p></div>
<div class="card"><h3>D30</h3><p>完成 3 件成果、3 筆採用證據與決策：續行、暫停、換 owner 或升級系統。</p></div>
</div>
</section>
<section class="card">
<h2>Purpose-to-Purpose E2E</h2>
<div class="grid flow">
<div class="step"><strong>原始目的</strong>讓 AI 落地助手變成可營收、可導入、可證明的產品系統。</div>
<div class="step"><strong>產出物</strong>30 天作戰台、資料模型、API、scorecard、people sync。</div>
<div class="step"><strong>人採用</strong>Owner 每日看板，產品線每週交 artifact，客戶回填使用證據。</div>
<div class="step"><strong>指標改善</strong>成果數、採用證據、逾期清除率、pipeline、營收。</div>
<div class="step"><strong>錢路徑</strong>三線產品組合包導向季度 1200 萬，保留能成交的線，砍掉空轉線。</div>
</div>
</section>
<section class="grid two">
<div class="card">
<h2>三線產品導入矩陣</h2>
<table>
<thead><tr><th>產品線</th><th>D7 artifact</th><th>D14 evidence</th><th>D30 decision</th></tr></thead>
<tbody>
<tr><td>AI Coach</td><td>Coach onboarding demo + rubric</td><td>1 位內部/客戶使用回填</td><td>續行或換場景</td></tr>
<tr><td>AgentERP</td><td>Lead-to-cash 或 source-to-pay workflow spec</td><td>1 條企業流程試跑證據</td><td>升級系統或暫停</td></tr>
<tr><td>FDE</td><td>FDE delivery playbook + 客戶 discovery 表</td><td>1 個客戶場景與付費假設</td><td>指定 owner 與 pipeline</td></tr>
</tbody>
</table>
</div>
<div class="card">
<h2>資料 / API / 權限</h2>
<p><strong>Tables:</strong> <code>ai_adoption_projects</code>, <code>product_workstreams</code>, <code>adoption_artifacts</code>, <code>usage_evidence</code>, <code>action_items</code>, <code>d30_decisions</code>.</p>
<p><strong>APIs:</strong> <code>POST /ai-adoption/workstreams</code>, <code>POST /ai-adoption/evidence</code>, <code>GET /ai-adoption/scorecard</code>, <code>POST /ai-adoption/d30-decision</code>.</p>
<p><strong>Permissions:</strong> owner 決策；product owner 上傳成果；customer-facing owner 回填證據；finance 看 pipeline/營收。</p>
</div>
</section>
<section class="grid three">
<div class="card"><h2>驗收</h2><p>D30 以前至少 3 件 openable artifact、3 筆採用/回填證據、80% 逾期待辦清除率、1 個明確決策。</p></div>
<div class="card"><h2>人的能力提升</h2><p>Owner 從「知道要導入 AI」升級成會判斷哪條線有營收、哪條線要停損、誰要負責。</p></div>
<div class="card"><h2>下一輪</h2><p>做成 PLS dashboard：自動拉 action items、artifact、LINE evidence、pipeline，產出 D30 decision pack。</p></div>
</section>
<section class="card source">
<h2>Market Maturity Inputs</h2>
<p>McKinsey 指出 agentic AI 正從實驗走向擴張，但多數企業仍卡在 pilot 到 scaled impact 的轉換：<a href="https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai">McKinsey State of AI 2025</a>.</p>
<p>UiPath 與 Deloitte 的 Agentic ERP 強調從 task automation 走向端到端流程 orchestration，對應 AgentERP 產品線：<a href="https://www.uipath.com/newsroom/uipath-expands-alliance-with-deloitte-for-agentic-erp">UiPath Agentic ERP</a>.</p>
<p>Epicor 2026 推出 AI-driven 90-day ERP go-live，顯示企業 AI/ERP 導入成熟標準正在壓縮到 90 天以內、強調 time-to-value：<a href="https://www.epicor.com/en-us/newsroom/news-releases/epicor-delivers-90-day-cloud-go-lives/">Epicor 90-day ERP</a>.</p>
</section>
</main>
</body>
</html>
Raw
artifact-url-or-pr.md

Artifact URL or PR

Primary artifact: https://gist.github.com/esz135888/387580ff96fad9da5d12fdf027f07fe6#file-ai-adoption-assistant-30day-command-center-html

Public Gist: https://gist.github.com/esz135888/387580ff96fad9da5d12fdf027f07fe6

Verification commands:

  • curl -I -L -s "https://gist.github.com/esz135888/387580ff96fad9da5d12fdf027f07fe6#file-ai-adoption-assistant-30day-command-center-html" | head -n 8
  • gh gist view 387580ff96fad9da5d12fdf027f07fe6 --files

Verification result: primary URL returned HTTP/2 200; public Gist includes 13 files.

Raw
data-model.md

Data Model

Tables

ai_adoption_projects

Field Type Notes
id uuid Primary key
parent_project_id uuid Mother project
north_star text Quarter revenue and integration target
d30_due_at date D30 decision date
owner_id uuid Louis or assigned owner
status enum active, paused, owner_change, system_upgrade, completed

product_workstreams

Field Type Notes
id uuid Primary key
adoption_project_id uuid Parent
product_line enum ai_coach, agent_erp, fde
owner_id uuid Product owner
d7_artifact_target text Expected artifact
d14_evidence_target text Expected evidence
revenue_hypothesis numeric Pipeline/quarter target
status enum not_started, building, validating, scaled, stopped

adoption_artifacts

Field Type Notes
id uuid Primary key
workstream_id uuid Related product line
artifact_type enum demo, workflow_spec, playbook, dashboard, pr, deployment
url text Openable artifact
owner_id uuid Uploader
accepted boolean Reviewer acceptance
created_at timestamptz Timestamp

usage_evidence

Field Type Notes
id uuid Primary key
workstream_id uuid Related product line
evidence_type enum line_reply, meeting_note, customer_trial, github_pr, drive_doc, payment
ref_url text Evidence link
signal_quality enum weak, medium, strong
nps integer Nullable
willingness_to_pay numeric Nullable
created_at timestamptz Timestamp

action_items

Field Type Notes
id uuid Primary key
workstream_id uuid Nullable
owner_id uuid Assignee
title text Action
due_at date Due
status enum open, blocked, done, overdue
evidence_required boolean Default true

d30_decisions

Field Type Notes
id uuid Primary key
adoption_project_id uuid Parent
decision enum continue, pause, change_owner, upgrade_system
reason text Required
evidence_summary text Required
decided_by uuid Owner
decided_at timestamptz Timestamp

API / Sync

  • POST /ai-adoption/workstreams: create product line.
  • POST /ai-adoption/artifacts: add openable artifact.
  • POST /ai-adoption/evidence: attach usage/owner evidence.
  • GET /ai-adoption/scorecard: current D1/D7/D14/D30 progress.
  • POST /ai-adoption/d30-decision: record final decision.
  • POST /ai-adoption/sync/action-items: sync PLS action items and overdue status.

Permissions

  • Owner: decide D30 and assign product owners.
  • Product owner: create workstream artifacts and action items.
  • Customer-facing owner: attach evidence and pipeline values.
  • Finance: view revenue pipeline and actual bookings.
  • PLS worker: sync action items and calculate scorecard.

Audit / Rollback

Every artifact, evidence and D30 decision writes audit_events. If an artifact is invalid, mark accepted=false, preserve URL, and require replacement rather than deleting history.

Raw
decision-record.md

Decision Record

Decision

Use system / project / scorecard / agent as the solution shape for AI落地助手導入 30 天推進.

Options Considered

  1. Status memo only

    • Pros: Fast.
    • Cons: Does not force artifacts, evidence or D30 decision.
    • Decision: Rejected.

  2. Project checklist only

    • Pros: Clear tasks.
    • Cons: Does not connect product lines, revenue path, evidence and owner decision.
    • Decision: Insufficient.

  3. 30-day adoption command center

    • Pros: Connects AI Coach, AgentERP, FDE, artifacts, evidence, action items, revenue and D30 decision.
    • Cons: Requires daily operating discipline.
    • Decision: Recommended.

Recommendation

Run the project as a 30-day adoption operating system. If a product line cannot produce D7 artifact and D14 evidence, it must be paused or reassigned instead of staying open.

Adoption Status

Ready for D1 owner assignment and D7 artifact build.

Feedback If Not Adopted

If the team rejects this cadence, the project should be paused or reduced to one product line until ownership and commercial target are explicit.

Raw
e2e-verification.md

E2E Verification

Plan

  1. Publish primary HTML and appendices to public Gist.
  2. Verify primary URL returns HTTP 200.
  3. Verify Gist includes 13 files.
  4. Upload files to PLS deliverable id.
  5. Complete with stable public artifact URLs.

Primary Artifact

https://gist.github.com/esz135888/387580ff96fad9da5d12fdf027f07fe6#file-ai-adoption-assistant-30day-command-center-html

Evidence

  • Published public Gist: https://gist.github.com/esz135888/387580ff96fad9da5d12fdf027f07fe6
  • Verification command: curl -I -L -s "https://gist.github.com/esz135888/387580ff96fad9da5d12fdf027f07fe6#file-ai-adoption-assistant-30day-command-center-html" | head -n 8
  • File list command: gh gist view 387580ff96fad9da5d12fdf027f07fe6 --files
  • Result: primary URL returned HTTP/2 200; Gist file list showed all 13 files.

Acceptance Mapping

  • Openable main artifact: ai-adoption-assistant-30day-command-center.html.
  • Owner/due/acceptance: production-brief.md.
  • Data/toolbox path: data-model.md, production-readiness.md.
  • E2E proof: this file.
  • Decision record: decision-record.md.
Raw
learning-memory.json
{
"project": "AI落地助手導入 30 天推進",
"market_learning": [
"Enterprise AI adoption is moving from pilots to workflow-scale impact, but most organizations still struggle to operationalize value.",
"Agentic ERP maturity emphasizes end-to-end process orchestration and governance.",
"Implementation time-to-value is compressing toward 90 days or less, so PLS should use 30-day proof points rather than open-ended planning."
],
"pls_next_checks": [
"Does each product line have an owner and a D7 openable artifact?",
"Does each product line have real usage evidence or owner reply by D14?",
"Is revenue pipeline attached to product-line decisions?",
"Is D30 decision recorded with evidence rather than left implicit?"
],
"assumptions": [
"AI Coach, AgentERP and FDE can each produce one artifact within D7.",
"At least one line can produce usage evidence by D14.",
"Quarter revenue target can be allocated across product lines."
],
"next_iteration": "Build live PLS dashboard that syncs action items, artifact URLs, LINE evidence and pipeline values into D30 decision pack."
}
Raw
market-maturity.md

Market Maturity

Sources

  1. McKinsey State of AI 2025

  2. UiPath and Deloitte Agentic ERP

  3. Epicor 90-day AI-driven ERP go-live

PLS Gap

The project already has signals and a North Star, but lacks a hard evidence loop: artifacts, owner replies, usage proof, pipeline value and D30 decision.

This Round Closes

  • Defines 30-day operating command center.
  • Adds three product line matrix.
  • Adds data model and API.
  • Adds acceptance tests and D30 decision rules.
  • Adds people sync and learning memory.

Remaining Gap

Implement live PLS dashboard that syncs action items, artifact links, LINE evidence and revenue pipeline.

Raw
people-sync.md

People Sync

Target People

  • Louis: D30 decision owner and revenue target owner.
  • AI Coach owner: D7 demo and D14 usage evidence.
  • AgentERP owner: workflow spec and enterprise process pilot.
  • FDE owner: delivery playbook and customer discovery evidence.
  • Finance/ops: pipeline and bookings evidence.

LINE Draft

AI落地助手導入這輪已整理成 30 天作戰台：AI Coach、AgentERP、FDE 三條線都要在 D7 前各交一件可打開成果，D14 前至少拿到一筆真實使用/回填證據，D30 做「續行 / 暫停 / 換 owner / 升級系統」決策。

Louis 只需要先拍板三件事：三條線的 owner、D7 要交的 artifact、季度 1200 萬營收要由哪條線主扛。

如果 D7 沒有 artifact 或 D14 沒有 evidence，不再延長空轉，直接進暫停或換 owner。

Escalation

若 48 小時內沒有 owner 回覆，PLS 應把三線縮成一條最可能成交的 AgentERP/FDE 組合線，避免三線同時空轉。

Raw
production-brief.md

Production Brief

場景

專案：AI 推進專案：AI落地助手導入 的 30 天推進。母專案 North Star：30 天內完成 AI Coach、AgentERP、FDE 三大核心產品整合，達成季營收 1200 萬元。

本輪把逾期訊號改造成 production operating pack：三線產品導入矩陣、資料模型、權限、API、驗收、people sync、D30 決策機制。

D1 / D7 / D14 / D30

  • D1: 建立三線產品矩陣、owner、客戶場景、資料欄位、第一批 action items。
  • D7: 每條線至少 1 件可打開成果：AI Coach demo、AgentERP workflow、FDE playbook。
  • D14: 取得 1 次真實使用/回填證據，開始測 NPS、願付費率、pipeline 金額。
  • D30: 完成 3 件成果、3 筆採用證據與決策：續行、暫停、換 owner 或升級系統。

Purpose-to-Purpose E2E

原始目的：讓 AI 落地助手變成可營收、可導入、可證明的產品系統。

產出物：30 天導入作戰台、資料模型、API、scorecard、people sync、learning memory、decision record。

人採用：Owner 每日看板，產品線每週交 artifact，客戶/內部回填使用證據。

價值/錢路徑：三線產品組合包導向季度 1200 萬；保留能成交與能交付的線，停掉空轉線。

Owner / Due / Acceptance

  • Owner: Louis.
  • Product owners: AI Coach owner, AgentERP owner, FDE owner.
  • Due: D7 三件 artifacts；D30 三件成果 + 三筆使用證據 + D30 決策。
  • Acceptance: 可驗收成果數 >= 3、owner 回覆/採用證據 >= 3、逾期待辦清除率 >= 80%、D30 決策完成。
Raw
production-readiness.md

Production Readiness

Ready

  • Openable primary HTML command center.
  • D1/D7/D14/D30 path.
  • Product line matrix for AI Coach, AgentERP, FDE.
  • Data model and API spec.
  • Acceptance tests.
  • People sync.
  • Decision record.
  • Market maturity references.

Not Yet Deployed

  • No live PLS dashboard module was deployed.
  • No actual LINE or revenue pipeline sync was executed.
  • No real owner reply evidence has been received in this round.

Production Path

  1. Create workstream rows for AI Coach, AgentERP and FDE.
  2. Assign product owners.
  3. Attach D7 artifact URLs.
  4. Sync LINE/meeting/GitHub/Drive usage evidence.
  5. Connect pipeline value and willingness-to-pay fields.
  6. Generate D30 decision pack automatically.

Rollback / Stop Rule

If by D14 there is no real usage evidence and no owner response, pause or change owner. If a product line has evidence but no revenue path, keep as internal capability but remove from 1200 萬 revenue plan.

Raw
skill-usage.md

Skill Usage

Selected Skills / Tools

  • using-superpowers: session skill discipline was already checked in this thread.
  • Web search: used to validate current enterprise AI/agentic ERP adoption maturity and time-to-value practices.
  • apply_patch: used for file creation.
  • GitHub Gist: used for stable openable primary artifact.
  • curl / gh gist view: used for verification.
  • PLS helper: used for doctor, touch, claim, context/progress retry, upload-files and complete.

Evidence

Note

Context/progress helper initially returned fetch failed; claim payload contained enough contract data to build first, with context retried before upload.

Raw
solution-selection.md

Solution Selection

Selected Types

  • system
  • project
  • scorecard
  • agent
  • governance

Why This Combination

AI落地助手導入 is not a single deliverable. It has three product lines, a revenue target, adoption evidence requirements, action item clearance and a D30 decision. The smallest viable shape is a system-backed project scorecard with worker sync.

Why Not Smaller

A one-page document or communication script would not create artifacts, usage evidence, or a D30 decision record.

Why Not Larger

A fully deployed PLS module is a next-round implementation. This round creates the production pack, data model, API and acceptance path that make it buildable.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment