Work Memo Skill

This skill provides intelligent work information recording with multi-dimensional tracking capabilities.

Purpose

Record work-related information (tasks, meetings, calls, etc.) with automatic categorization, prioritization, and organization. This is a record-only system - it does not execute tasks.

When to Use

Use this skill when:

• User wants to record work information
• User mentions recording tasks, meetings, calls, or work events
• User provides URLs to articles/documentation to save for later
• User wants to track work items with priorities and contexts
• User needs to organize work by tags, contexts, or dates

Capabilities

1. Natural Language Processing

• Parse work descriptions from natural language
• Extract work type, priority, tags, contexts, ands
• Detect urgency and importance levels automatically
• Support both English and Chinese

2. Multi-Dimensional Tracking

• Type: task, meeting, call, email, bugfix, feature, review, research, coding, design
• Urgency: 1-5 scale (detected from keywords like "urgent", "紧急")
• Importance: 1-5 scale (detected from keywords like "important", "重要")
• Tags: #work, #urgent#meeting, etc.
• Contexts: @office, @home, @phone, @computer, etc.
• Dates: today, tomorrow, specific dates

3. Eisenhower Matrix Classification

Automatic quadrant assignment:

• Q1: Urgent + Important (Do First)
• Q2: Not Urgent + Important (Schedule)
• Q3: Urgent + Not Important (Delegate)
• Q4: Not Urgent + Not Important (Eliminate)

4. URL Content Processing

• age content from URLs
• Extract and summarize main content
• Store URL as context for easy reference
• Auto-tag with "web" and "reading"
• Default type: "research"

Implementation

For Text Input

1. Parse Input: Extract work details from natural language

   • Identify title/description
   • Detect work type from keywords
   • Find priority indicators (urgent, important)
   • Extract tags (#tag) and contexts (@context)
   • Parse date references

2. Execute Recording: Use skills/scripts/memo_command.py

   python3 skills/scripts/memo_command.py "<work description>"

3. Return Confirmation: Include:

   • Recorded title
   • Type
   • Urgency/importance levels
   • Eisenhower quadrant
   • Tags and contexts
   • Date

For URL Input

date URL**: Check format (http:// or https://)

2. Fetch Content: Use web fetching capabilities to retrieve webpage

3. Summarize: Extract title and create summary

4. Record: Store with:

   • Webpage title as record title
   • URL in contexts
   • Summary in description
   • Type: research
   • Tags: web, reading

5. Return Confirmation: Include URL and summary
   ples

Text Input

Input: 紧急会议今天 #work @office
Output: {
  "status": "recorded",
  "title": "紧急会议",
  "type": "meeting",
  "urgency": 4,
  "importance": 3,
  "eisenhower_quadrant": "Q1",
  "tags": ["work"],
  "contexts": ["office"],
  "date": "2026-02-05"
}

URL Input

Input: https://example.com/article
Output: {
  "status": "success",
  "title": "Article Title",
  "url": "https://example.com/article",
  "summary": "Content summary...",
  "type": "research",
  "tags": ["web", "reading"],
  "eisenhower": "Q2"
}

Natural Language Syntax

Tags

• Format: #tagname
• Examples: #work, #urgent, #project, #meeting, #bug

Contexts

• Format: @context
• Examples: @office, @home, @phone, @computer

Priority Keywords

• Urgency: urgent, emergency, asap, 紧急, 非常紧急
• Importance: important, critical, key, 重要, 关键

Dates

• Natural: today, tomorrow, this week, 今天, 明天, 本周
• Specific: 2026-02-05

Output Format

Always return a JSON-formatted confirmation with:

• status: "recorded" or "success"
• title: Work item title
• type: Work type
• urgency: 1-5 scale
• importance: 1-5 scale
• eisenhower_quadrant or eisenhower: Q1-Q4
• tags: Array of tags
• contexts: Array of contexts
• date or created_at: Date timestamp

Storage System

Records are stored as Markdown files with YAML frontmatter:

• Location: ~/work-memo/records/
• Structure: YYYY-MM/DD/[id]_[title].md
• Format: YAML frontmatter + Markdown content
• Benefits:
  • Human-readable and editable
  • Version control friendly
  • Easy backup and sync
  • No database required
  • Portable across systems

Example file:

~/work-memo/records/2026-02/05/a1b2c3d4_urgent_meeting.md

Notes

• This is a record-only system - no execution happens
• All data is stored as Markdown files (not database)
• Natural language parsing provides best results
• More contextetter automatic categorization
• URLs require internet connection
• The /memo command uses this skill automatically
• Records can be manually edited in any text editor
• Compatible with git for version control

Scripts

• Main command: skills/scripts/memo_command.py
• Query parser: skills/scripts/query_parser.py
• Web scraper: skills/scripts/web_scra
• Reporting: skills/scripts/reporting.py
• Schema: skills/scripts/schema.py