weekly-review · AI/ML · Mesrai Skills

Quick Start

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

thoughts/shared/pm/reports/ - This week's plan (what you intended)
thoughts/shared/pm/reports/ - Daily plans from this week (what actually happened)
thoughts/shared/pm/prds/ - PRDs modified this week
thoughts/shared/product/meeting-notes/ - Meeting notes from past 7 days
thoughts/shared/pm/launches/ - Launches that happened this week
thoughts/shared/product/decisions/ - Decisions made this week
thoughts/shared/pm/analyses/ - Research conducted
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

Calculate week to review:
- Default: Current week (if Friday or later)
- If Monday-Thursday: Ask "Review last week or current week?"
- If user specified: Use that week
Check if review already exists:
- Look for thoughts/shared/pm/reports/YYYY-WXX-weekly-review.md
- 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:

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

Template:

---
week: YYYY-WXX
week_start: YYYY-MM-DD
week_end: YYYY-MM-DD
quarter: Q[X] YYYY
---

# Weekly Review - Week of [Month] [DD], [YYYY]

## TL;DR

- **PRDs:** [X active], [Y advanced], [Z stalled]
- **Launches:** [N features shipped]
- **Meetings:** [M total], [P key decisions]
- **Completion rate:** [X%] of planned tasks done
- **Key win:** [Biggest accomplishment]
- **Key challenge:** [Biggest blocker/lesson]

---

## Strategic Progress

**Quarter Goal:** [Primary goal for Q]
**Progress This Week:** [What moved forward]

| Goal     | Start of Week | End of Week | This Week | Status      |
| -------- | ------------- | ----------- | --------- | ----------- |
| [Goal 1] | X%            | Y%          | +Z%       | ✅ On track |
| [Goal 2] | A%            | B%          | +C%       | ⚠️ Behind   |

**Velocity check:**

- [x] weeks left in quarter
- [Y%] progress needed per week to hit goal
- [Z%] actual progress this week
- **Assessment:** [On track / Need to accelerate / Ahead]

---

## Top 3 Priorities Review

[Compare planned vs actual]

### Priority 1: [Title]

**Planned:** [What we intended to achieve]
**Actual:** [What we achieved]

**Tasks:**

- ✅ [Task 1] - Done
- ✅ [Task 2] - Done
- [ ] [Task 3] - Carried over because [reason]

**Status:** ✅ Complete / 🟡 Partial / ❌ Not started

**Key outcome:**

- [What this unlocked or enabled]

**Learning:**

- [What went well or what to change]

---

### Priority 2: [Title]

[Same structure]

---

### Priority 3: [Title]

[Same structure]

---

## PRD Pipeline

| PRD    | Stage (Start of Week) | Stage (End of Week) | Movement    | Next Action       |
| ------ | --------------------- | ------------------- | ----------- | ----------------- |
| [Name] | Team Kickoff          | Planning Review     | ✅ Advanced | Get eng estimates |
| [Name] | Solution Review       | Solution Review     | ⚠️ Stalled  | Need legal review |
| [Name] | -                     | Team Kickoff        | 🆕 New      | Scope and plan    |

**Analysis:**

- **Advanced:** [PRD X] moved forward because [stakeholder signed off / design done / etc.]
- **Stalled:** [PRD Y] blocked on [dependency / decision / resource]
- **Recommendation:** [What to prioritize next week to unblock]

---

## Launches & Impact

### Shipped This Week

[If anything launched]

**[Feature Name]** (Launched [Day])

**Success Criteria (from PRD):**

- [Metric 1]: Target [X], Actual [Y] ([+/-Z%])
- [Metric 2]: Target [A], Actual [B] ([+/-C%])

**Early assessment:** ✅ On track / ⚠️ Needs attention / ❌ Below target / 🚀 Exceeding

**Insights:**

- [User feedback received]
- [Unexpected behavior observed]
- [Next iteration needed]

---

### Post-Launch Monitoring

[Features launched in past 4 weeks still being monitored]

| Feature | Launch Date | Key Metric | Target | Actual | Trend | Status   |
| ------- | ----------- | ---------- | ------ | ------ | ----- | -------- |
| [Name]  | [Date]      | [Metric]   | [X]    | [Y]    | [↗↘→] | [✅⚠️❌] |

---

## Key Decisions Made

1. **[Decision]** ([Date] - [Meeting])
   - **Context:** [Why this came up]
   - **Decision:** [What was decided]
   - **Rationale:** [Why we chose this]
   - **Owner:** [Who's executing]
   - **Impact:** [What this affects]
   - **Doc:** [Link if exists]

2. **[Decision 2]**
   [Same structure]

---

## Meetings & Stakeholder Pulse

### Meetings This Week: [Total]

| Day | Meeting | Attendees | Outcome              | Value     |
| --- | ------- | --------- | -------------------- | --------- |
| Mon | [Topic] | [Names]   | [Decision/Alignment] | 🟢 High   |
| Tue | [Topic] | [Names]   | [Info sharing]       | 🟡 Medium |
| Wed | [Topic] | [Names]   | [Cancelled]          | ⚫ None   |

**Meeting load:** [X] hours / 40 = [Y%]
**Deep work time:** [Z] hours (vs [A] hours planned)

**Value assessment:**

- 🟢 High value: Led to decision or unblocked work
- 🟡 Medium value: Useful context but no immediate action
- 🔴 Low value: Could have been async or skipped

**Recommendation:** [Which meetings to keep/change/cancel]

---

### Stakeholder Pulse

**High engagement this week:**

- **[Name]** - [Why: Multiple syncs, key decision, strong collaboration]
  - Impact: [What this enabled]
  - Continue: [Keep this cadence / Increase collaboration]

**Needs attention:**

- **[Name]** - [Why: Haven't synced in 2+ weeks, blocking issue, misalignment suspected]
  - Impact: [What's at risk]
  - Action: [Specific next step - schedule 1:1, send update, etc.]

**New relationships:**

- **[Name]** - [Met for first time, context]
  - Follow-up: [Add to stakeholder profiles, schedule regular sync]

---

## User Research & Insights

[Only include if research happened]

**New Research This Week:**

- **[Interview/Study]** - [Date]
  - Key finding: [Insight]
  - Validates: [Which hypothesis or PRD]
  - Challenges: [What assumption or approach]

**Recurring Themes:**

- **[Theme 1]** - Mentioned in [X] sources
  - Evidence: [Quote or data point]
  - Implication: [What this means for roadmap]

- **[Theme 2]** - Validates hypothesis from [PRD]
  - Evidence: [Quote or data point]
  - Recommendation: [Accelerate this PRD / Pivot approach]

**Competitive Intelligence:**
[If competitive analysis updated]

- [Competitor] launched [Feature]
- Implication: [How this affects our strategy]

---

## Tasks & Execution

**Completion Metrics:**

- **Completed:** [X] tasks
- **Carried over:** [Y] tasks ([Z%] carry-over rate)
- **Added mid-week:** [A] tasks (scope creep indicator)

**By initiative:**

### [Initiative/PRD Name]

- ✅ [Task completed]
- ✅ [Task completed]
- [ ] [Task carried over] - **Why:** [Blocked by X / Deprioritized for Y / Under-estimated]

### [Initiative 2]

[Same structure]

**Patterns:**

- Tasks that took longer than expected: [Type/category]
- Blockers that repeated: [Dependency on X person/team]
- Tasks that got bumped repeatedly: [Maybe not actually important?]

---

## Learnings & Patterns

**What Worked Well:**

- **[Approach/Decision]** - [Why it was effective]
  - Example: [Specific instance]
  - Repeat: [How to apply this pattern again]

**What Didn't Work:**

- **[Mistake/Inefficiency]** - [What happened]
  - Impact: [Consequence]
  - Root cause: [Why this happened]
  - Fix: [Specific change for next time]

**Process Improvements:**

- [ ] [Specific improvement to implement]
  - Why: [Problem it solves]
  - How: [Concrete action]
  - Owner: You
  - By when: [Next week / Next sprint]

**Personal Development:**
[If applicable]

- Skill practiced: [What you worked on]
- Feedback received: [From whom, about what]
- Growth area identified: [What to develop]

---

## Next Week Preview

### Top 3 Priorities (Draft)

[Based on this week's outcomes, suggest next week's priorities]

1. **[Priority 1]** - [Why: Carries over from this week / New urgent item / Strategic next step]
2. **[Priority 2]** - [Why]
3. **[Priority 3]** - [Why]

> Note: Run `/weekly-plan` to formalize these and add detail

---

### Key Meetings Next Week

[From calendar if available]

- **[Day]:** [Meeting] - [Goal/Outcome needed]
- **[Day]:** [Meeting] - [Prep needed]

---

### Items to Unblock

| Item       | Blocked Since | Blocked By                   | Action Needed  |
| ---------- | ------------- | ---------------------------- | -------------- |
| [PRD/Task] | [Date/Week]   | [Person/Dependency/Decision] | [Specific ask] |

**Priority unblocks:**

1. [Most critical blocker to address Monday]
2. [Second priority]

---

## Metrics to Monitor Next Week

[Features to keep watching]

- **[Feature 1]** - [Why: Early launch / Trending down / Critical metric]
  - Watch: [Specific metric]
  - Check: [Daily / Every other day]
  - Flag if: [Threshold or condition]

---

_Generated: [Timestamp]_
_Data sources: [Weekly plan, Daily plans, PRDs, Meeting notes, Linear, Analytics]_
_Next: Run `/weekly-plan` to plan next week_

Step 5: Follow-Up Prompts

After generating review, prompt user with contextual suggestions:

Always offer:

"Week synthesized and saved! Next steps:

Plan next week? Run /weekly-plan (5-10 min) - I've drafted initial priorities above

Share with team? I can format this as a stakeholder update

What would help?"

If significant wins:

"🎉 Nice work on [Achievement]! Worth documenting this:

Add to portfolio/resume

Share in team update

Capture as case study for future reference

Want me to help with any of these?"

If patterns emerged:

"📊 I noticed some patterns:

[Recurring blocker X] appeared [Y] times

[Task type Z] consistently took 2x longer than estimated

Want to dig into these and create process improvements?"

If learnings captured:

"💡 This week's learnings worth remembering:

[Learning 1]

[Learning 2]

I'll surface these in future planning. Want me to add to thoughts/shared/pm/context/lessons-learned.md?"

If metrics concerning:

"⚠️ [Feature] metrics need attention:

[Metric] is [X%] below target

Trending [down/flat] since launch

Run /feature-results for deeper analysis? Or schedule stakeholder review?"

Integration with Other Skills

Before /weekly-review:

/daily-plan - Ran throughout the week (provides daily context)
/meeting-notes - Captured meeting outcomes
/prd-draft - Created/updated PRDs this week

After /weekly-review:

/weekly-plan - Plan next week based on this review
/decision-doc - Document key decisions made
/status-update - Share with stakeholders
/feature-results - Deep dive on launched features

Parallel use:

/impact-sizing - Validate completed work had expected impact
/competitor-analysis - If competitive intel emerged this week

Tips for Best Results

When to run:

Best time: Friday afternoon (4-5pm)
- Week is fresh in memory
- Can plan next week immediately after
- Creates clean mental closure for weekend
Alternative: Monday morning (reflect before planning)
Avoid: Mid-week (incomplete picture)

What makes a good review:

✅ Honest about what didn't go well (not just wins)
✅ Specific about patterns (not vague "work harder")
✅ Actionable improvements (concrete next steps)
✅ Connects to strategy (not just task completion)
❌ Just a task list (misses the "why" and learnings)
❌ All problems, no wins (demotivating)

How to build the habit:

Week 1-2: I'll prompt you Friday afternoon
Week 3-4: You'll start expecting it (ritual forming)
Week 5+: Feels incomplete without it

Use the output:

Reference in 1:1s with manager (shows progress)
Share with stakeholders (transparency)
Compare month-over-month (velocity trends)
Review quarterly (pattern detection across multiple weeks)

Related Skills

Before this:

/daily-plan - Daily execution throughout week
/weekly-plan - Set priorities at start of week
/meeting-notes - Captured throughout week

After this:

/weekly-plan - Plan next week immediately after review
/status-update - Share summary with stakeholders
/decision-doc - Formalize key decisions made

Periodic use:

/feature-results - Monthly deep dive on launched features
/quarter-review - (If exists) Quarterly synthesis of weekly reviews

Output Quality Self-Check

Before delivering the weekly review, verify:

Plan vs. actual compared: If a weekly plan existed, every planned priority is addressed with a status (complete, partial, not started) and a reason for any gap.
Learnings are specific and actionable: Each learning includes what happened, why, and a concrete change for next time. "Work harder" is not a learning.
Next week priorities are drafted: At least 3 draft priorities for next week are suggested, grounded in this week's outcomes and strategic goals.
Blockers have owners: Every unresolved blocker has a specific action and person to contact on Monday.
Metrics referenced where available: If launches happened or metrics data exists, actual numbers are cited (not just "things went well").
Appropriate length: Default review is ~150 lines. Full review is longer but still organized with clear section headers. Do not generate a full review unless the user asked for one.
Honest about what did not go well: The review includes at least one thing that did not go as planned, with root cause analysis. A review with only wins is incomplete.