AgentSkillsCN

summarize

对文本、文章及对话进行摘要提炼

SKILL.md
--- frontmatter
name: summarize
description: Summarize text, articles, and conversations
version: 1.0.0
author: ClawdBot
tags: [summarization, text-processing]
requires_bins: []
requires_env: []
requires_config: []

Text Summarization

Summarize long texts, articles, web pages, and conversations.

Available Tools

  • •web_fetch: Fetch article content
  • •read_file: Read local files
  • •sessions_history: Get conversation history

Summarization Strategies

1. Extract Key Points

  • •Identify main ideas
  • •Bullet point format
  • •Focus on actionable items

2. Executive Summary

  • •2-3 paragraph overview
  • •Include context and conclusions
  • •Professional tone

3. TL;DR

  • •Single paragraph
  • •Casual tone
  • •Essential information only

Usage Examples

User: "Summarize this article: {URL}"

  1. •Fetch article with web_fetch
  2. •Extract main content
  3. •Generate summary with key points

User: "Give me a TL;DR of our conversation"

  1. •Get conversation history with sessions_history
  2. •Identify key topics and decisions
  3. •Create concise summary

User: "Summarize this document"

  1. •Read file content
  2. •Identify structure (sections, headings)
  3. •Summarize each section
  4. •Provide overall summary

Best Practices

  • •Ask user for preferred format (bullets, paragraph, TL;DR)
  • •Include word count if requested
  • •Preserve critical details (names, dates, numbers)
  • •Maintain original tone for technical content
  • •Highlight action items in meeting summaries

Output Formats

Bullet Points

code
Key Points:
- Point 1
- Point 2
- Point 3

Executive Summary

code
Summary:
{2-3 paragraphs covering main themes}

Key Takeaways:
- Takeaway 1
- Takeaway 2

Meeting Summary

code
Meeting Summary
Date: {date}
Participants: {names}

Discussion:
- Topic 1
- Topic 2

Decisions:
- Decision 1
- Decision 2

Action Items:
- [ ] Task 1 - Owner
- [ ] Task 2 - Owner