agentleFS
guang-lee-cnguang-lee-cn/scaffold-skills/skills/report/SKILL.md

report

Skill2 starsChanged 13 days agoMITView on GitHub

技术调研/研究类报告的**呈现层**规范:把研究结论组织成研发总监可读、研发可落地、审计可追溯的正式报告。当需要输出研究报告、技术调研报告、选型报告、版本对比报告、迁移可行性报告时使用。本 skill 管「怎么呈现」,研究过程(怎么调研)归 research/。触发词:调研报告/研究报告/对比报告/迁移评估/报告呈现/写报告/整理结论成文。

93 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use opportunities, export and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • ansvisorSkill

    opportunities

    Finds content gaps, prompts worth tracking, and the off-site sources AI cites instead of the brand.

    Also in 3 projects that use this

  • soilmassSkill

    export

    Animation export configurations including format selection, resolution, framerate, quality, and optimization settings. Use when rendering animations for web, mobile, video, or social media. Provides tokens for all export parameters across Lottie, GIF, video, and sprite sheet formats.

    Also in 4 projects that use this

  • GoogleSkill

    quality

    Evaluates whether a GitHub issue is spam, empty, needs more information, or is OK to proceed.

    Also in 3 projects that use this

  • genomewalkerSkill

    resume

    Resume a thread by loading its ~800-token context capsule

    Also in 7 projects that use this

  • tokyubevoxelverseSkill

    checkup

    Give a repository a full health exam — doc rot, dependency staleness, CI decay, git hygiene, dead links, docs contradicting code — and grade it. Use when the user wants to know what quietly rotted, or a prioritized maintenance list.

    Also in 3 projects that use this

  • rajnavakotiSkill

    scan

    Run the AI dashboard scanner and generate the HTML report.

    Also in 5 projects that use this

The file

Files
SKILL.md

report · 报告呈现层规范 v1.0

配套:research/(研究执行层:8 接口研究法 + 审计钩子)+ 本文(呈现层)。 原则:一次报告四层齐备——缺 L1 结论不可读,缺 L2 不可审计,缺 L3 不可信,缺 L4 没人读完。

一、定位澄清(与 research/ 的分工)

层归属回答什么
研究执行层research/怎么调研(8 接口、快研、研究计划)
报告呈现层本 skill怎么交付(四层规范:结构/追溯/防幻觉/可读性)

接口:研究过程中按 research/ 审计钩子产出三张表(数据源登记/证据清单/数字登记),报告期直接引用,不回炉重查。

二、四层正交规范

L1 结构层:信息在哪

  1. BLUF 段:主文档第 1 段出结论——核心结论 + 推荐动作 + 时间/代价 + 不行动风险,3-8 句。读者只看这段可决策。
  2. 受众分层(深度不同,事实与严重性不变):
    受众前置框架体量
    研发总监项目风险 + 决策点 + 代价前置段 ≤1 页
    研发落地技术方案 + 范围清单 + 依赖 + 改动面正文 L1 章节
    审计追溯 + 证据 + 复现命令附录
  3. 评分量表(0-40,32+ 可交付):结论位置(0/5/10)、独立成文、可行动性(推荐+主责+日期)、受众契合。

L2 追溯层:结论凭什么信

  1. 验证徽章:[TV] 工具验证 / [HV] 人工复核 / [LA] 模型推断待复核。每条结论挂徽章,未复核默认 [LA]。
  2. 置信度分级:GROUNDED(直接证据)/ INFERRED(推导待验证)/ SPECULATIVE(需人工确认)/ ABSENT(未分析,显式缺口)。
  3. 四层抽象追溯(面向总监章节):L0 技术事实(原文,进附录)→ L1 影响 → L2 项目风险 → L3 行动项;每层可回溯,不可追溯的结论不写。
  4. 防造:未知写 TBD,估算标 [ESTIMATED],训练知识标 [FROM TRAINING - verify independently];不编造趋势与对比;无基线全部标 NEW。

L3 防幻觉层:数字哪来的

  1. 数字锁:写正文前先建数字表(数值 + 口径 + 出处 + verified 位)。正文数字以编号引用;引用了表里没有的条目 = 不通过。口径写不成一句完整话的数字禁用。
  2. 规则三级迁移:能写函数/一行判据 → 进校验脚本;能计数 → 自检报数;机器判不了 → 才留本手册。
  3. 渲染后逐页检查:编号/目录/页码机械检查;版心溢出/图表截断/孤悬标题逐页人眼(12 页拼一表扫全)。自检假警报比漏报贵——宁可少报不可天天误报。

L4 可读性层:人读不读得下去

  1. 句法:单句(含逗号构件)≤30 字,>40 字不可接受;一句一意;拆欧化长句("之所以…是因为…"类);优先主动语态、具体动词(write/close/matched),少用虚动词。
  2. 术语:专有名词/FastDDS 术语继承源头命名不机翻(Requester/Replier/ContentFilteredTopic/QoS/TypeObject/SampleIdentity);首次出现给一句人话定义,后续只用源名;代码符号/变量/文件名原样。
  3. 段落节奏:一段一主题、中心句段首;段长 50-200 字,>250 必拆;长句后接短句,结论句独立成行;列表项一句话;数字与结论相邻不隔段。
  4. 人味:只写方案事实,禁解释性废话/心理活动/官腔("本报告旨在…");先给"这意味着什么",再给细节。

三、报告骨架(默认)

§0 元信息(任务/范围/版本对/日期/审计说明)
§1 BLUF 段(总监 3 分钟读完,全挂徽章)
§2 版本基线
§3 规范层对照
§4 声明层对照(CHANGELOG/文档/README/TODO)
§5 落地层证据(diff/源码)
§6 交叉矩阵(规范×声明×落地)
§7 研发落地评估(迁移可行性/改动面/依赖)
§8 结论(≤1 页)
§9 审计附录(数据源表/复现命令/证据清单/数字表)

四、反模式(四层汇总死线)

  • L1:时间线摘要(那是 abstract);埋没推荐;不同受众共用一份
  • L2:无来源结论;徽章缺失;为受众降严重性
  • L3:正文现编数字;口径不清的数字;手册写机器能判的规则
  • L4:30 字以上长句连排;术语中英混译;无中心句段落;数字与结论分离

五、交付前检查(四层逐项)

  • L1:BLUF 前三句出结论;受众分层;评分 ≥32
  • L2:每条结论有徽章;置信度分级完整;追溯附录与正文一一对应
  • L3:数字全部可回数字表;口径成句;校验脚本 0 报错
  • L4:抽样 10 句均 ≤30 字;术语无中英混译;段落有中心句;全文无官腔
  • 渲染:页码/编号一致,逐页人眼过完

六、与外部 skill 的关系

  • 本规范为整合产物,来源:pm-claude-skills/executive-summary(L1)、dtsong/executive-brief-skill(L2)、alchaincyf/huashu-report(L3)、Google Technical Writing + 中文技术文档写作风格指南 + 用户写作偏好(L4)。三源原文证据落盘调研项目 2026-10-24/sources/skills/。

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.