Docs & Knowledge · nexu-io/html-anything

exec-briefing-memo

Name: exec-briefing-memo
Author: nexu-io/html-anything

Decision needed + recommendation + evidence + tradeoffs, compressing complex materials into an actionable one-page memo ready for executive decision-making.

L1 Basic · Source registration onlyV1 Commit-PinnedDual-pass Audit

Author nexu-ioGitHub nexu-io/html-anything ↗Stars 8827License Apache-2.0（仓库根 LICENSE = Apache License 2.0；GitHub API license.spdx_id = Apache-2.0；next/package.json 亦声明 Apache-2.0）Commit 553ed98c28

Control levels describe skill capabilities and suggested control measures; they do not represent quality or safety certifications, nor do they dictate usability. Usage decisions rest entirely with users and administrators. Different agent hosts apply varying constraints; do not assume hosts provide automatic protection. Third-party skill copyrights belong to their original authors. Use at your own risk.

1How It Works & Mechanism

This skill is not an executable program, but rather "one directory = one template": the frontmatter of SKILL.md is used for display in the selector, while the body serves as the prompt sent to the model; adding a template equals adding a directory, requiring no modifications to TypeScript.

next/src/lib/templates/loader.ts

routes rescan disk and the client refetches `/api/templates`.

The header comments in the same file specify the directory convention: SKILL.md (frontmatter + prompt body) / example.md (optional input sample) / example.html (optional preview). This skill directory precisely adheres to this structure.

The actual dispatched prompt = platform shared design directives (SHARED_DESIGN_DIRECTIVES) + the SKILL.md body of this skill + user content, assembled into a single prompt via assemblePrompt.

next/src/lib/templates/shared.ts

Wrap a per-template instruction body with the shared design directives and

Therefore, the body of this skill is only responsible for the "layout / style / component library" section; output disciplines (single-file HTML, strict use of real user data, content-driven quantity) originate from the shared directives, so template authors do not need to redeclare them.

The final document bypasses the file system: the shared directives strictly forbid file tools such as Write/Edit/Bash, requiring the model to stream the entire HTML directly as the response body (stdout is captured and written to disk by the front end).

next/src/lib/templates/shared.ts

**禁止使用 Write / Edit / MultiEdit / Bash / Create / 任何文件系统工具**。不要把 HTML 写到任何 \`.html\` 文件里。前端直接捕获你的 stdout 文本, 文件落盘由前端负责。

This is the mechanism behind the "zero disk writes from the skill itself": the host front end writes the file, not the skill; there are also no scripts whatsoever in this directory.

The deliverable is defined as "an actionable one-pager ready for decision in 3 minutes" rather than minutes, weekly reports, or PRDs, enforcing a mandatory 9-section structure (memo header / decision needed / recommendation / why now / key facts / tradeoff table / risks / decision path / next actions).

next/src/lib/templates/skills/exec-briefing-memo/SKILL.md

【意图】这不是会议纪要、不是周报、不是 PRD。它的唯一目标是帮助决策者在 3 分钟内理解问题并拍板。

The structure contains mandatory criteria: recommendations must be explicit with a confidence level, strictly disallowing phrases like "can be considered"; Tradeoffs must compare upside / cost / risk / reversibility across Option A/B/C; each item in Risks must provide a mitigation action.

Above-the-fold discipline is embedded into the design requirements: conclusions and recommendations must appear first, and background must never lead.

next/src/lib/templates/skills/exec-briefing-memo/SKILL.md

- 首屏必须直接呈现 decision + recommendation, 不要先铺陈背景。

The sample implements this directly: within the first section, header pills (Executive memo / Owner / Deadline / Audience) appear first, followed immediately by the prominent "Decision needed" heading and the green Recommendation card on the right (Approve a narrow Enterprise Preview. + Confidence: medium-high + Reversible).

This template is the only one among the nine that comes with an optional style assets directory: three ready-to-use HTML styles under assets/ (board-memo default / decision-command dark command deck / board-paper board proposal), strictly requiring one to be selected per scenario with no mixing allowed.

next/src/lib/templates/skills/exec-briefing-memo/SKILL.md

- `assets/board-memo.html`: 默认风格。浅色高管 memo, 适合 CEO/CFO/CRO、运营、产品决策。

The same section provides selection rules: default to board-memo if unspecified; use decision-command if materials emphasize urgency/risk/operational command; use board-paper for board presentations/formal approvals. All three assets are Tailwind CDN single files.

The sample output shares the exact same source as assets/board-memo.html: example.html in the directory is an instantiated version of the "default style" asset.

next/src/lib/templates/skills/exec-briefing-memo/assets/board-memo.html

<div class="label">Tradeoff table</div>

A diff of both files shows no difference (byte-for-byte identical); therefore, what is seen in the preview is the board-memo style itself, while the other two styles (decision-command dark, board-paper serif) only provide base styling templates in assets without bundled sample outputs.

Content authenticity and evidence gaps are codified into the rules: fabricating numbers, customers, budgets, or dates is forbidden; when key information is missing, list Evidence gaps while still providing a provisional recommendation.

next/src/lib/templates/skills/exec-briefing-memo/SKILL.md

- 如果缺少关键信息, 在 Evidence gaps 中列出, 但仍给出基于现有证据的 provisional recommendation。

The accompanying example.md serves as an input sample for "fragmented materials → one-page memo": mixed signals regarding sales, product, financials, risks, and CXO concerns for FlowLedger, used to verify whether the model can synthesize them into a decision frame.

The decision path is productized into three side-by-side cards (Approve / Reject / Need evidence), rather than a paragraph of "next steps" text.

next/src/lib/templates/skills/exec-briefing-memo/example.html

<h3 class="font-extrabold">Need evidence</h3>

This corresponds to SKILL.md rule 8: "next steps for each of the three paths: approve / reject / ask for more evidence"; the Tradeoff table and Risks & mitigations in the sample are also standalone sections, forming an actionable decision-making chain: "conclusion — evidence — tradeoffs — risks — paths".

The server-side /api/convert performs only three tasks: loads the skill from disk by templateId, assembles the prompt, spawns the local agent CLI, and forwards the output via SSE stream.

next/src/app/api/convert/route.ts

prompt = assemblePrompt({ body: skill.body, content, format });

The same file first executes `const skill = loadSkill(templateId);` (returns 400 if the template does not exist), and then passes it to invokeAgent; the skill directory itself spawns no processes and has no network access.

2Core Capabilities

01Compress fragmented materials into a one-page decision memo (Decision needed + Recommendation + Confidence)

02Build a tradeoff table comparing Option A/B/C (upside / cost / risk / reversibility)

03Provide a 3-way decision path (approve / reject / ask for more evidence) with next steps for each

04Next actions checklist (owner / due date / expected artifact)

05Three optional style assets, select one per scenario (board-memo / decision-command / board-paper)

06Paired risk and mitigation outputs (mitigation actions for 3-5 risks each)

07Built-in input sample (fragmented materials on whether FlowLedger should launch an Enterprise Plan)

08Layout discipline where conclusions are placed above the fold

4Recommended Governance & Evidence

Suggested Controls for Safe Operation

For Agent Runtime

Can be auto-loaded

For Administrators

Evidence & Triggers (File Paths & Excerpts)

SKILL.md: Pure prompt instructions and rule text; contains no script execution or external impacts.

Audit Notes: Grading target = the skill directory itself (SKILL.md / example.html / example.md / three style assets): zero scripts (aside from the Tailwind CDN script in the generated HTML), zero credential access, zero disk-write directives, and no runtime network call code. According to the Batch 4 final rules: "Artifact references" are not graded—SKILL.md references assets/board-memo.html, assets/decision-command.html, assets/board-paper.html, and example.html, which in turn reference cdn.tailwindcss.com and fonts.googleapis.com; these endpoints are fetched by the final HTML in the client browser and are not called outbound during skill execution. As static style/font CDNs, they transmit no user data and require no credentials, thus remaining green. Verification of internal directory endpoints: ① https://cdn.tailwindcss.com ×3 (line 7 in each of the three assets); ② https://fonts.googleapis.com (board-memo/example imports Inter + Noto Sans SC, board-paper imports Source Serif 4, decision-command imports Inter). Host-side context provided for clarity: local agent CLI spawn, preview iframe allow-scripts + allow-same-origin. Contains no orange criteria (no external generation backend calls) and no red criteria (no reverse engineering / spoofing / bypass / downgrade).

5Second-Pass Verification

[discrepancy] 9 段强制结构中样例实际覆盖的部分 — 样例含 Decision needed / Recommendation（含 Confidence: medium-high）/ Tradeoff table / Risks & mitigations / Decision path（Approve·Reject·Need evidence）/ header（Owner·Deadline·Audience）；缺 Why now、独立 Key facts 块（仅 4 张指标卡）、Next actions、Evidence gaps。
[ok] example.html 与 assets/board-memo.html 字节相同 — diff 输出为空；挂载的预览即默认风格资产的实例化版本。
[ok] 三套风格资产存在且风格迥异 — board-memo（--paper:#faf9f5）、decision-command（--bg:#101418 深色）、board-paper（--paper:#fbfaf6 + Source Serif 4 + 双线 rule）三份齐全。
[discrepancy] README 收录情况 — 仓库 README 的模板表（105 处模板链接）未列出 exec-briefing-memo，grep briefing/memo 只命中 deck-safety-alert；属文档滞后——目录存在、loader 能加载、featured=8 已声明。
[ok] 成稿版式依赖 Tailwind 运行时 — 三份资产均 <script src="https://cdn.tailwindcss.com"></script> 且大量使用 Tailwind 工具类（p-7 / lg:grid-cols-[1.15fr_.85fr] 等）；离线时栅格与间距会失效，仅内联 <style> 部分保留。
[ok] 「不要捏造数字」与样例输入的关系 — example.md 提供了 18 inbound / 7 accounts / $18k-35k / 4-6w 等素材，样例成稿的 4 张指标卡（Enterprise asks 18、Large accounts 7、Potential ACV $18k-35k、Roadmap cost 4-6w）逐项对得上，未被编造。
[ok] 目录内无脚本、无凭证读取 — 除 Tailwind CDN script 外无 JS；无 token/env 读取；assets 只被 SKILL.md 指定为风格参考，不执行。

6Verdict & Scenarios

A template solely focused on decision-making: structural clauses are all decision criteria (explicit recommendation + confidence, three paths, four tradeoff axes), naturally rendering outputs actionable.

Above-the-fold conclusions tailored to executive 3-minute reading habits: decision + recommendation highlighted cards sit at the very top, followed by evidence and tradeoffs.

Style as rules: three assets + explicit selection criteria (default board-memo / urgent decision-command / board-level board-paper), offering far greater control than ad-hoc model styling.

Anti-hallucination discipline: requires listing Evidence gaps when information is missing while still offering a provisional recommendation, balancing honesty with actionability.

Includes an input sample with a clear implementation path: messy CXO materials → one-page memo, ready for direct comparative learning.

Best for: Ideal for compressing meeting notes, research findings, and sales/financial feedback into an actionable one-page memo; suitable for product and operations leads who provide decision inputs to CEOs, CFOs, boards, or investors; suitable as a standardized vehicle for go/no-go decisions, budget approvals, and release gates.

Not for: Not suitable for lengthy reports requiring full argumentative processes (this template intentionally compresses to one page); not suitable for meeting minutes, weekly reports, or PRDs (explicitly excluded by SKILL.md); not suitable for external marketing collateral (lacks narrative and brand expression); not suitable for financial analyses requiring multi-page charts due to heavy data volumes (metric cards + tables have limited capacity).

Install Agent-Ready

① Verified Mirror 2026-09-15

Option A · Direct DownloadDownload exec-briefing-memo.tar.gz

sha256: 69bd73088bbba9e0…

Option B · Copy Guide for Agent

install.json

Agents automatically download and verify mirrors from this JSON guide.

② Upstream GitHub

GitHub Repo ↗

Pinned to commit 553ed98c28.