Report Outline

v1.0.0

在有了研究大纲后使用。将研究大纲转化为详细的报告结构和执行计划,明确每个章节的内容、数据来源、分析方法。

0· 115·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for mingginwan/data-report-outline.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Report Outline" (mingginwan/data-report-outline) from ClawHub.
Skill page: https://clawhub.ai/mingginwan/data-report-outline
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install data-report-outline

ClawHub CLI

Package manager switcher

npx clawhub@latest install data-report-outline
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
The name and description match the SKILL.md: it transforms a research outline into a report structure and execution plan. It requests no binaries, credentials, or config paths — all proportional to an instruction-only outlining tool.
Instruction Scope
Instructions stay focused on report structure, chapter-level breakdown, dependencies, and checks. The SKILL.md asks the agent to 'save the report structure' with a filename pattern and to announce when the skill is used; these are reasonable but implementation-dependent (the agent runtime will determine where the file is written). It references a separate 'research skill' for splitting multi-topic outlines — composition is expected but means full behavior depends on other skills present. The instructions do not ask to read arbitrary system files, access environment variables, or transmit data to external endpoints.
Install Mechanism
No install spec (instruction-only). Nothing is downloaded or written to disk by the skill itself beyond the expectation that the agent may save an output file; this is low-risk.
Credentials
No environment variables, credentials, or config paths are requested. The skill does not ask for secrets or unrelated service tokens.
Persistence & Privilege
always is false and the skill does not request persistent elevation or alter other skills' configurations. Autonomous invocation is allowed by platform default but the skill itself does not request extra privileges.
Assessment
This skill appears coherent and low-risk: it contains only instructions for producing a detailed report plan and requests no credentials or installs. Before installing/using, consider: 1) Confirm where the agent will save the generated file (the SKILL.md asks to 'save to an appropriate location' but doesn't specify a path or storage policy). Make sure that destination is acceptable and access-controlled. 2) If your agent composes this with a separate 'research skill', verify that skill's permissions and behavior. 3) Review outputs for accidental inclusion of sensitive data from your prompt or environment before distributing. 4) If you enable autonomous agent runs, remember that the skill may cause the agent to spawn follow-up tasks (e.g., parallel research agents) — review those agents' scopes. If you want a stricter assessment, provide the agent runtime's save-path behavior and any linked 'research' skill definitions; with that I can reassess any additional risks.

Like a lobster shell, security has layers — review code before you run it.

latestvk975e3mw8gwrjm1cs87vqa1c0n83prvr
115downloads
0stars
1versions
Updated 1mo ago
v1.0.0
MIT-0

报告结构与执行计划

将研究大纲转化为可执行的报告结构。假设执行者对主题背景了解有限——明确写出每个章节要覆盖什么、数据从哪来、用什么分析方法、预期产出是什么。把整份报告拆解为可独立完成的模块。

启动时宣布: "正在使用 report-outline skill 制定报告结构。"

范围检查

如果研究大纲涉及多个独立主题,应在 research skill 阶段就拆分。如果没有拆分,建议分别制定报告结构——每份报告应该是一个独立、完整的分析。

报告结构模板

每份报告计划必须以此开头:

# [报告标题] 结构计划

**目标:** [一句话描述这份报告要回答什么问题]

**受众:** [谁会读这份报告,他们的背景和关注点]

**核心结论预期:** [预期这份报告会得出什么类型的结论]

---

章节结构

### 章节 N:[章节名称]

**目的:** 这个章节要回答什么问题

**数据来源:**
- 来源 1:`具体文件/链接/API`
- 来源 2:`具体文件/链接/API`

**分析方法:**
- 具体的分析步骤(不是"分析数据",而是"对比 A 和 B 在 X 维度的差异")

**预期产出:**
- 表格/图表/结论的具体描述

**依赖:**
- 是否依赖其他章节的结果

**预估篇幅:** 约 X 字

不允许占位符

每个章节必须包含执行所需的具体内容。以下是计划失败的标志——永远不要写:

  • "待定"、"TODO"、"后续补充"
  • "进行适当分析"(什么分析?)
  • "收集相关数据"(什么数据?从哪里?)
  • "参考章节 N"(重复写出来——执行者可能不按顺序读)
  • 描述要做什么但不说明怎么做的步骤

执行顺序

标注章节间的依赖关系,给出推荐的执行顺序:

## 执行顺序

1. 章节 2(独立,无依赖)→ 章节 3(独立)→ 章节 1(依赖 2、3 的结论)→ 章节 4(总结)

可并行:章节 2 和章节 3
必须串行:章节 1 在 2、3 之后

自查

写完报告结构后,用新鲜眼光对照研究大纲检查:

  1. 大纲覆盖: 研究大纲里的每个要点,都能在报告结构中找到对应章节吗?列出遗漏。
  2. 占位符扫描: 搜索上面"不允许占位符"中列出的模式。修掉。
  3. 一致性检查: 后面章节引用的数据/结论,和前面章节定义的一致吗?
  4. 受众匹配: 分析深度和术语使用,适合目标受众吗?

发现问题直接修改。如果大纲有未覆盖的要点,补上对应章节。

保存与交接

保存报告结构到合适位置,文件名格式:YYYY-MM-DD-<主题>-报告结构.md

完成后提供执行选择:

"报告结构已保存到 <路径>。两种执行方式:

1. 并行研究(推荐) — 对独立章节同时启动研究代理,加快速度

2. 逐章撰写 — 按顺序逐章完成,每章完成后审阅

选哪种?"

Comments

Loading comments...