askill
beauty-step1

beauty-step1Safety 95Repository

Document content analysis and merging. Automatically invoked during step 1 of the beauty command to fully understand source document content, extract key information, and establish content structure. 文档内容分析合并。在beauty命令的步骤1执行时自动调用,用于完整理解源文档内容,提取关键信息,建立内容结构。

1 stars
1.2k downloads
Updated 2/6/2026

Package Files

Loading files...
SKILL.md

Beauty 步骤1:文档内容分析合并 / Beauty Step 1: Document Content Analysis

目标 / Goal

完整理解源文档内容,提取关键信息,建立内容结构。

Fully understand source document content, extract key information, and establish content structure.

⚠️ Token限制处理机制 / Token Limit Handling Mechanism

如果文档过长无法一次性读取完整 / If document too long to read at once:

正确做法(使用继续机制)/ Correct Approach (Use Continue Mechanism):

步骤1.1:读取文档前半部分 / Step 1.1: Read first half of document
├─ 使用 Read 工具读取文档(offset: 0, limit: 500)
   Use Read tool to read document (offset: 0, limit: 500)
├─ 提取前半部分的章节结构
   Extract chapter structure from first half
├─ 记录前半部分的数据点
   Record data points from first half
└─ 输出:"步骤1.1完成 - 已分析文档前半部分(0-500行)
       请输入'继续'以分析后半部分"
   Output: "Step 1.1 complete - Analyzed first half of document (lines 0-500)
           Please input 'continue' to analyze the second half"

【等待用户输入"继续" / Wait for user to input "continue"】

步骤1.2:读取文档后半部分 / Step 1.2: Read second half of document
├─ 使用 Read 工具读取文档(offset: 500, limit: 500)
   Use Read tool to read document (offset: 500, limit: 500)
├─ 提取后半部分的章节结构
   Extract chapter structure from second half
├─ 记录后半部分的数据点
   Record data points from second half
└─ 输出:"步骤1.2完成 - 已分析文档后半部分(500-1000行)
       文档分析100%完成"
   Output: "Step 1.2 complete - Analyzed second half of document (lines 500-1000)
           Document analysis 100% complete"

【如果文档超过1000行,继续分片读取 / If document exceeds 1000 lines, continue segmented reading】

步骤1.3:合并分析结果 / Step 1.3: Merge analysis results
├─ 整合所有部分的分析结果
   Integrate analysis results from all parts
├─ 生成完整的章节结构清单
   Generate complete chapter structure list
├─ 生成完整的数据点清单
   Generate complete data point list
└─ 进入步骤2
   Proceed to Step 2

❌ 禁止做法(偷工减料)/ Prohibited Actions (Cutting Corners):

错误做法1:只读前几行 / Wrong Approach 1: Only read first few lines
├─ Read 文件(limit: 50)
   Read file (limit: 50)
└─ ❌ 跳过了后面的内容
   ❌ Skipped remaining content

错误做法2:使用摘要 / Wrong Approach 2: Use summaries
├─ 请用户总结文档内容
   Ask user to summarize document content
└─ ❌ 可能丢失细节
   ❌ May lose details

错误做法3:分步执行但不说明 / Wrong Approach 3: Execute in steps without explanation
├─ 只生成前半部分内容
   Only generate first half of content
└─ ❌ 用户不知道要继续
   ❌ User doesn't know to continue

执行要求 / Execution Requirements

  • ✅ 完整阅读源文档,不做任何修改或删减 Fully read source document without any modification or deletion

  • ✅ 识别文档类型(报告、分析、方案、研究等) Identify document type (report, analysis, plan, research, etc.)

  • ✅ 提取核心内容元素 / Extract core content elements:

    • 标题层次结构(H1 → H2 → H3) Title hierarchy (H1 → H2 → H3)
    • 关键数据和数值 Key data and numerical values
    • 主要结论和见解 Main conclusions and insights
    • 重要建议和行动项 Important recommendations and action items
    • 对比关系和表格数据 Comparison relationships and table data
  • ✅ 建立内容逻辑结构 / Establish content logical structure:

    • 主题分组 Topic grouping
    • 因果关系 Causal relationships
    • 时间顺序 Chronological order
    • 优先级排序 Priority ranking
  • ✅ 记录所有定量数据(数字、百分比、货币等) Record all quantitative data (numbers, percentages, currency, etc.)

输出产物 / Output Artifacts

  • 内容结构大纲(包含所有章节和要点) Content structure outline (includes all chapters and key points)

  • 数据点清单(所有可用于可视化的数值) Data point list (all values available for visualization)

  • 关键结论列表(必须完整保留) Key conclusions list (must be fully preserved)

验证标准 / Validation Criteria

  • 所有原文内容已提取 / All original content extracted
  • 无内容丢失或遗漏 / No content loss or omission
  • 数据点完整记录 / Data points fully recorded
  • 逻辑结构清晰 / Logical structure clear

完成后输出 / Output Upon Completion

✅ 步骤1:文档内容分析合并 - 100%完成
✅ Step 1: Document Content Analysis - 100% Complete

分析摘要 / Analysis Summary:
- 文档类型 / Document type: [报告/分析/方案/研究 / Report/Analysis/Plan/Research]
- 章节数量 / Chapter count: [N个H2章节 / N H2 chapters]
- 数据点数量 / Data point count: [N个 / N points]
- 关键结论 / Key conclusions: [N个 / N conclusions]

输出产物 / Output Artifacts:
✓ 内容结构大纲 / Content structure outline
✓ 数据点清单 / Data point list
✓ 关键结论列表 / Key conclusions list

验证结果 / Validation Results:
✓ 内容完整性:100% / Content integrity: 100%
✓ 数据准确性:已验证 / Data accuracy: Verified
✓ 结构清晰度:优秀 / Structure clarity: Excellent

准备进入步骤2 / Ready to proceed to Step 2

Install

Download ZIP
Requires askill CLI v1.0+

AI Quality Score

92/100Analyzed 2/11/2026

An exceptionally well-structured, bilingual skill document that provides a robust framework for chunked document analysis. It includes precise tool usage instructions (offset/limit), explicit 'prohibited actions' to ensure data integrity, and clear validation criteria.

95
100
65
95
98

Metadata

Licenseunknown
Version-
Updated2/6/2026
PublisherWithin-7

Tags

No tags yet.