/week - Weekly Review

You are helping the user with their weekly review. Read the last 7 daily notes and generate a summary.

Behavior

1. Read daily notes from 99_System/Daily Notes/ for the last 7 days
2. Read active workstreams from 00_Dashboard/Active Workstreams.md
3. Read Goals.md from 00_Dashboard/Goals.md for weekly targets
4. Aggregate and analyze:
   • Group activities by project/workstream
   • Calculate averages for energy, mood, sleep
   • Count job outreach attempts, applications, habits
   • Compare actuals vs weekly targets from Goals.md
   • Calculate energy allocation % by goal
   • Note any patterns (low energy days, avoided tasks, etc.)
5. Trigger pattern detection engine on last 4 weeks of data
   • Update Goals.md "Pattern Detection" section with new insights
   • Refresh optimization suggestions based on latest data
6. Generate weekly review with Goal Progress section first
7. Offer to save as a weekly review file

Output Format

## WXX Review (Date Range)

### Goal Progress

**Read from Goals.md and show weekly targets vs actuals with visual indicators**

| Goal | Target | Actual | Status |
|------|--------|--------|--------|
| Applications | 8-12/week | X | 🚨⚠️✅🔥 |
| Outreach | 15-25/week | Y | 🚨⚠️✅🔥 |
| PM Interviews | 2/week | Z | 🚨⚠️✅🔥 |
| Exercise | 5x/week | A | 🚨⚠️✅🔥 |
| No drinks Mon-Thu | 4x/week | B | 🚨⚠️✅🔥 |

**Energy Allocation:**
- PM: X% (target: 60%)
- Gig: Y% (target: 30%)
- SaaS: Z% (parked — passive scanning only)
- Energy: A% (target: 10%)

**Top 2 Optimization Suggestions from Goals.md:**
1. [Type] Suggestion text
2. [Type] Suggestion text

---

### Projects
- Project Name: X sessions, key accomplishments
- Job Search: X/15 outreaches [status indicator]

### Practices
- AI Learning: X hours (topics covered)
- Exercise: X days
- Other notable activities

### Patterns
- Energy: avg X.X (note any dips)
- Mood: avg X.X
- Sleep: avg X.X hrs
- Vices: summary

### What Worked
- (positive patterns observed)

### What Didn't
- (things avoided, friction points)

### This Week's Learnings
- Summary of items processed from explore queue this week
- Key insights added to Domain MOCs or KB files
- Cross-cutting patterns or connections discovered

### Suggested Next Week
1. Specific actionable suggestion
2. Another suggestion
3. Third suggestion

[If monthly boundary - add AI re-evaluation section here]

Save as WXX review?

Status Indicators

Use status indicators from _shared/pattern-detection.md (🚨⚠️✅🔥 based on % of target). Celebrate wins explicitly.

File Location

Weekly reviews go in: 99_System/Weekly Reviews/YYYY-WXX.md

Use ISO week numbers.

Weekly Review Template (when saving)

---
Domain: "[[D - Career & Job Search]]"
tags:
  - weekly-review
  - productivity
created: YYYY-MM-DD
week: YYYY-WXX
---
Nav: [[YYYY-WXX|< Prev Week]] | [[YYYY-WXX|Next Week >]]
Related: [[Active Workstreams]] | [[Home]]
Daily Notes: [[YYYY-MM-DD]] | [[YYYY-MM-DD]] | ... (all 7 days)

# Week XX Review

[Generated content here]

---

## Next Week's Focus
1.
2.
3.

Linking Rules

When creating weekly reviews:

1. Nav links: Include prev/next week links (only if those files exist)
2. Daily Notes line: Link to all daily notes from the week being reviewed
3. Project links: Link to all projects mentioned in the summary:
   • Use display aliases: [[Legata Index|Legata]], [[P - Job Search Strategic Project (Full Version)|Job Search]]
4. In the body: When referencing specific accomplishments, link to the relevant project
5. Domain links: If a domain got significant attention, link to it in the summary

Inbox Processing

After generating the weekly summary:

Universal Inbox Check

1. Check universal inbox: Read 01_Inbox/00_INBOX.md and count items in Inbox section
2. If >5 items: "You have X items in inbox. Triage now? (y/n)"
3. If yes: Offer /triage all or process interactively
4. Don't force it - keep review flow smooth, user can triage later

Explore Queue Processing

1. Check explore queue: Read 01_Inbox/00_EXPLORE.md and count unprocessed items
2. Offer batch processing: "You have X unprocessed explore items. Process all now? (y/n)"
3. If yes: Process all remaining items using /explore skill behavior
   • Fetch, summarize, categorize, and append each item
   • Handle failures gracefully (move to Failed or Manual Review sections)
4. Generate learning digest: Summarize all items processed this week
   • Highlight new learnings added
   • Identify patterns across topics
   • Suggest one action to apply next week

Next Week Plan Generation

After inbox and explore processing, offer to generate next week's plan:

1. Offer: "Generate W{N+1} plan? (y/n)"
2. If yes:
   • Read 02_Projects/Job Search/Scout + Dossier/R - Outreach Log.md → find all rows with follow-up dates in next week's date range
   • Read 00_Dashboard/Goals.md → confirm current Goal 1/2 priorities and weekly targets
   • Generate theme-day file using this structure:
     • Mon: Follow-up sweep (overdue + stragglers from this week) + any unfinished non-negotiables carried forward
     • Tue: Applications batch (90 min) + Scout run + Goal 2 action block (30 min)
     • Wed: Outreach batch from Tuesday's scout run + Goal 2 research/define block
     • Thu: Applications batch (90 min) + hot reply follow-ups
     • Fri: Close loop — scholarship/admin check, outreach log cleanup, W{N+2} queue prep
   • Populate Non-Negotiables from Goals.md weekly targets (applications count, outreach count, goal 2 action)
   • Pre-populate Follow-up queue table with outreach log rows whose follow-up dates fall in next week
   • Save as 99_System/Weekly Reviews/YYYY-WXX-Plan.md
3. If no: skip — next week's plan can be generated manually or skipped

Monthly AI Boundary Re-evaluation

Every ~4 weeks (when ISO week number % 4 ≈ 0), add this section to the output after "Suggested Next Week":

### 🔄 Monthly AI Boundary Re-evaluation

It's been about a month - time to re-test AI capabilities:

**Tasks you currently do manually:** [Scan recent daily notes and list 2-3 recurring manual tasks]

**Action this week:** Pick one task above and try delegating it to AI. Models improve every few weeks - capabilities you tested last month may work now.

**Remember:** Your mental model of "what AI can do" falls behind reality within weeks. Regular re-testing prevents leaving value on the table.

Detection logic: Check if week_number % 4 == 0 OR if it's been 4+ weeks since last monthly check (track in a comment or simple heuristic).

Analysis Guidelines

• Be honest about underperformance but not preachy
• Job outreach target is 15/week - measure against this
• Look for correlations (low sleep = low energy days?)
• Note any workstream that got zero attention
• Identify the "one thing" that would make next week better
• Include learning digest from explore processing

Conversation Style

• Present the summary first, then ask if they want to save
• If they want to adjust next week's focus, help them refine
• Keep it action-oriented, not guilt-inducing

Post-Invocation

After completing weekly review, always suggest context clear before next task. /week is high token cost - starting fresh maximizes remaining session capacity.