gather-week-context

name	gather-week-context
description	Collect and synthesize all daily archives for a given week. Returns aggregated data including wins, insights, people interactions, and completeness status.
disable-model-invocation	true
allowed-tools	Read, Bash(ls, date)

Gather Week Context Sub-Skill

Collects all daily archives for a specified week and synthesizes them into structured data for weekly review.

Input Arguments

Arguments are passed as key-value pairs:

• vault: Path to the vault
• week: Target week in YYYY-Www format
• week_start: First day of week (YYYY-MM-DD, Monday)
• week_end: Last day of week (YYYY-MM-DD, Sunday)

Instructions

1. Determine Workdays

Calculate all workdays (Monday-Friday) for the target week using the week_start date:

# week_start is always Monday
# Calculate: Monday, Tuesday, Wednesday, Thursday, Friday
# week_start + 0, +1, +2, +3, +4 days

2. Check Daily Archive Existence

For each workday, check if archive exists:

ls "$VAULT/00_Brain/Periodic/Daily/{date}.md" 2>/dev/null

Track:

• existing_archives[] - Dates with archives
• missing_days[] - Dates without archives (format: "YYYY-MM-DD (DayName)")

3. Read Each Daily Archive

For each existing archive, read and parse:

Frontmatter:

• date - The archive date
• day - Day of week
• energy - Energy level (High/Medium/Low)
• location - Work location
• focus_hours - Hours of focused work
• meetings - Number of meetings

Sections:

• priorities[] - The 3 priorities with completion status
• leadership_intention - The intention for the day
• meetings[] - Meeting blocks (extract person from 1:1s)
• wins - Personal, organisational, strategic subsections
• insights - what_went_well, what_could_be_better, key_insight
• carry_forward[] - Items marked for next day

4. Aggregate Data

Compile across all daily archives:

Metrics:

• total_meetings: Sum of all meeting counts
• total_focus_hours: Sum of all focus hours
• avg_energy: Average energy level (High=3, Medium=2, Low=1)
• energy_trend: "increasing", "stable", or "decreasing" based on Mon→Fri
• priority_completion_rate: Percentage of priorities marked complete

Collections:

• wins.personal[]: All personal wins with dates
• wins.organisational[]: All org wins with dates
• wins.strategic[]: All strategic wins with dates
• insights[]: All key insights with dates
• what_went_well[]: All positive patterns with dates
• what_could_be_better[]: All friction points with dates
• carry_forward[]: All incomplete items from final day

Entity Aggregates:

• unique_people[]: All [[PersonName]] references with interaction counts
• unique_projects[]: All [[project-name]] references with mention counts
• leadership_intentions[]: All daily intentions with dates

People Interactions: For each unique person, compile:

• interaction_count: How many days they appeared
• topics: Extracted from meeting notes and wins
• notable_moment: Most significant interaction (from wins/insights)

5. Generate Daily Links

Create wikilinks to each daily archive for the weekly archive:

- [[00_Brain/Periodic/Daily/2026-02-09|Monday]] - [One-line summary from day]
- [[00_Brain/Periodic/Daily/2026-02-10|Tuesday]] - [One-line summary]
...

Use the first line of daily notes or a summary of wins as the one-liner.

6. Return Structured Result

Success (full week):

{
  "success": true,
  "week": "2026-W07",
  "daily_archives": [
    {
      "date": "2026-02-09",
      "day_name": "Monday",
      "exists": true,
      "path": "/path/to/.../Periodic/Daily/2026-02-09.md",
      "energy": "High",
      "meeting_count": 5,
      "focus_hours": 3,
      "priority_completion": "2/3",
      "leadership_intention": "Focused",
      "wins_count": 4,
      "key_insight": "Delegation creates capacity"
        
    
    
       
       
       
       
       
       
       
       
       
       
       
        
    
  
   
   
     
     
     
     
     
     
           
           
    
     
           
           
    
     
           
           
    
  
   
     
           
           
    
     
           
    
     
           
           
    
  
   
         
         
  
   
         
         
  
   
         
  
   
    
       
       
         
         
       
    
    
       
       
        
        
       
    
  
   
         
         
  
   

Success (partial week):

{
  "success": true,
  "partial": true,
  "week": "2026-W07",
  "daily_archives": [...],
  "missing_days": ["2026-02-11 (Wednesday)", "2026-02-13 (Friday)"],
  "message": "3 of 5 workdays archived. Proceeding with available data.",
  "aggregates": {...},
  "wins": {...},
  "insights": [...],
  "people_interactions": [...],
  "carry_forward": [...],
  "daily_links" 

No archives found:

{
  "success": false,
  "error": "no_archives",
  "week": "2026-W07",
  "missing_days": ["2026-02-09 (Monday)", "2026-02-10 (Tuesday)", "2026-02-11 (Wednesday)", "2026-02-12 (Thursday)", "2026-02-13 (Friday)"],
  "message": "No daily archives found for this week.",
  "suggestion": "Run `/daily-review` for each day, or proceed with Week.md content only."
}

For backwards compatibility, also output human-readable summary:

Daily Archives for 2026-W07

Found: 5/5 workdays archived

Totals:
- Meetings: 18
- Focus Hours: 15h
- Energy: 2.4 avg (stable)
- Priority Completion: 73%

People: SarahK (3), MarcusT (2)
Projects: platform-migration (5), q2-roadmap (2)

Key Insights:
- Mon: "Delegation creates capacity"
- Tue: "Deep work compounds"
...

Design for Reuse

This sub-skill is designed to be reusable by:

• weekly-review: Primary consumer for weekly synthesis
• monthly-review: Can call multiple times to gather all weeks in a month
• quarterly-review: Can aggregate weekly data for quarterly patterns

The date range inputs allow flexibility for different time periods.