weekly-review

Quick Start

1. Run /weekly-review on Friday afternoon (best time) or Monday morning
2. I will scan your workspace: weekly plans, daily plans, PRDs, meeting notes, decisions, and launches from the past 7 days
3. I will generate a focused review comparing plan vs. actual, surfacing key wins, blockers, and learnings
4. Output goes to thoughts/shared/pm/reports/YYYY-WXX-weekly-review.md
5. After the review, I will suggest running /weekly-plan to plan next week

Default output is focused (~150 lines max). Say "full review" if you want the expanded version with stakeholder pulse, task-level execution metrics, and pattern analysis.

Purpose

End-of-week synthesis reviewing what you accomplished, what you learned, and what needs attention. Feeds into next week's planning and builds institutional memory.

Usage

• /weekly-review - Review current/past week
• /weekly-review last-week - Review previous week (if you forgot)

Context Routing

Check these files first:

1. thoughts/shared/pm/reports/ - This week's plan (what you intended)
2. thoughts/shared/pm/reports/ - Daily plans from this week (what actually happened)
3. thoughts/shared/pm/prds/ - PRDs modified this week
4. thoughts/shared/product/meeting-notes/ - Meeting notes from past 7 days
5. thoughts/shared/pm/launches/ - Launches that happened this week
6. thoughts/shared/product/decisions/ - Decisions made this week
7. thoughts/shared/pm/analyses/ - Research conducted
8. thoughts/shared/pm/frameworks/ - Quarter goals (to track progress)

MCP Queries (if available):

• Linear MCP - Tasks completed this week
• Analytics MCP - Metrics for features launched recently
• GitHub MCP - Code activity (if relevant to your role)
• Slack MCP - Key conversations and decisions

Fallback: File-based analysis of PM OS workspace + manual input for completions.

Workflow

Step 1: Determine Review Period

1. Calculate week to review:
2. Default: Current week (if Friday or later)
3. If Monday-Thursday: Ask "Review last week or current week?"
4. If user specified: Use that week

1. Check if review already exists:
2. Look for thoughts/shared/pm/reports/YYYY-WXX-weekly-review.md
3. If exists: Ask "Update existing review or create new version?"

Step 2: Data Collection

A. Weekly Plan (What Was Intended):

Read thoughts/shared/pm/reports/YYYY-WXX-weekly-plan.md:

Extract:

• Top 3 priorities for the week
• Key tasks under each priority
• Success criteria
• Expected meeting load

If no weekly plan exists:

• Note: "Week wasn't planned. Reviewing what happened only."
• Suggest: "Next week, run /weekly-plan on Monday for better focus."

B. PRD Progress:

Scan thoughts/shared/pm/prds/:

Method 1 - File modification dates:

# Files modified in the past 7 days
find thoughts/shared/pm/prds/ -name "*.md" -mtime -7

For each PRD touched this week:

• Read frontmatter or first section for current stage
• Compare to last week's stage (if weekly review from last week exists)
• Determine: Advanced, Stalled, or New

Method 2 - If Git available:

# PRDs with commits this week
git log --since="7 days ago" --name-only --pretty=format: | grep -E "prds/.*\.md$" | sort -u

Categorize:

• Advanced: Moved to next stage (Team Kickoff → Planning Review)
• Active: Work happened but didn't advance stage
• Stalled: No activity this week
• New: Started this week

C. Feature Launches:

Check thoughts/shared/pm/launches/:

• Launches completed this week
• Launch checklists finished
• Post-launch monitoring started

For each launch: If Analytics MCP available:

Query metrics since launch date
Compare to success criteria from PRD

Categorize:

• ✅ On track (meeting targets)
• ⚠️ Needs attention (below targets)
• ❌ Underperforming (significantly below)
• 🚀 Exceeding (beating targets)

D. Meetings & Decisions:

Scan thoughts/shared/product/meeting-notes/ from past 7 days:

For each meeting:

• Extract date, attendees, topic
• Look for: Decisions made, action items created, blockers identified

Check thoughts/shared/product/decisions/:

• Decision docs created this week
• Link to related meetings

Build stakeholder pulse:

• Who did you meet with most? (frequency)
• Who did you miss syncing with? (gaps)
• What topics dominated discussions? (themes)

E. Tasks Completed:

If Linear MCP available:

Query: Tasks completed in past 7 days
Group by: Priority, PRD/Initiative
Calculate: Planned vs actual completion rate

If MCP not available:

• Scan daily plans for checked-off tasks
• Scan meeting notes for completed action items

Categorize by initiative:

Initiative: [PRD Name]
- ✅ Task 1 (from Priority 1)
- ✅ Task 2 (from Priority 1)
- [ ] Task 3 (carried over - why?)

Calculate metrics:

• Tasks completed vs planned
• % completion rate
• Carry-over rate

F. User Research & Insights:

Check thoughts/shared/pm/:

• New interview notes this week
• Competitive analysis updates

Check thoughts/shared/pm/analyses/:

• Synthesis reports created
• Themes identified

Extract:

• Key findings
• Recurring themes (mentioned in multiple sources)
• Recommendations for roadmap

G. Learnings & Patterns:

This is where weekly review gets powerful - surfacing patterns.

From daily plans:

• What consistently took longer than expected?
• What got deprioritized every day? (maybe not important)
• What meeting prep was valuable vs not?

From outcomes:

• What decisions went well? (process to repeat)
• What decisions went poorly? (what to change)
• What blockers kept recurring? (systemic issue)

From stakeholder interactions:

• What communication worked well?
• What caused confusion or misalignment?
• Who needs more/less frequent updates?

Step 3: Analysis & Synthesis

PRD Pipeline Analysis:

For each PRD:

• Last week's stage → This week's stage
• Movement: ✅ Advanced / → Active / ⚠️ Stalled

Why analysis:

• Advanced: What unblocked it? (repeat this)
• Stalled: What's blocking? (action needed)

Strategic Alignment:

Read thoughts/shared/pm/frameworks/ for quarter goals.

For each goal:

• Which priorities/tasks contributed to it this week?
• Progress estimate: X% → Y% (did we move the needle?)
• Velocity: Are we on track for quarter target?

Pillar balance: If strategy has defined pillars:

• Pillar 1: X% of time this week
• Pillar 2: Y% of time
• Pillar 3: Z% of time

Compare to target allocation: Are we balanced?

Pattern Detection:

Look for:

• Recurring blockers: Same dependency/person blocked multiple things
• Underestimated tasks: Consistently took 2x longer than planned
• Overcommitted weeks: Planned 30 hours with 25 hours of meetings
• Meeting value: Which meetings led to outcomes vs were FYI only?
• Best working times: When did deep work happen? (protect these blocks)

Step 4: Generate Weekly Review

Create file: thoughts/shared/pm/reports/YYYY-WXX-weekly-review.md

Output Length Guidance:

Default (focused review, ~150 lines max). Include only:

1. TL;DR (5-6 bullet summary)
2. Priority Completion (plan vs actual for top 3 priorities)
3. Key Decisions Made (list with one-line rationale each)
4. Metrics Movement (table of metrics that changed)
5. Top 3 Learnings (what worked, what did not, what to change)
6. Next Week Preview (draft priorities + items to unblock)

Full review (when user asks for it). Expand to also include:

• Stakeholder pulse (engagement gaps, new relationships)
• Task-level execution metrics (completion rate, carry-over rate, scope creep indicator)
• PRD pipeline table with stage movement
• Meeting value assessment (high/medium/low for each meeting)
• Pattern analysis (recurring blockers, underestimated task types, best working times)
• User research and competitive intelligence updates

**Temp