Skip to content

Weekly-Review-Checklist.md

Latest commit

 

History

History
126 lines (85 loc) · 4.81 KB

Weekly-Review-Checklist.md

File metadata and controls

126 lines (85 loc) · 4.81 KB

Weekly Review Checklist

Purpose: A 15-minute ritual to keep your three authoritative files accurate, your AI collaboration healthy, and your project on track.

When to run: Once per week — Sunday evening or Monday morning recommended.
Time required: 10–20 minutes depending on project size.

Before You Start

Open these three files:

  • Strategy Master (strategy-master.md or equivalent)
  • Running Document (running-document.md or equivalent)
  • Canonical Numbers (canonical-numbers.md or equivalent)

Step 1: Numbers Audit (5 min)

Go through your Canonical Numbers file line by line.

  • Are all numbers still current? Update any that changed this week.
  • Are there any numbers marked TBD that you now have? Fill them in.
  • Did you use any numbers in work this week that aren't in the file? Add them now.
  • Are there any rows marked ⚠️ (unverified / needs follow-up)? Resolve or extend them.
  • Did you promise any data to someone (a client, a collaborator, a document) that isn't in the file? Add it with source.

Red flag: If you find a number in a deliverable you sent this week that isn't in the canonical file — log that in the failure log. It means a process broke down.

Step 2: Running Document Update (5 min)

  • Update "Last updated" date at the top.
  • Update "Current phase" if it changed.
  • Add any decisions made this week to the Decisions Log (even small ones).
  • Resolve any "Open Questions" that got answered this week. If still open, leave them.
  • Add any new open questions that emerged.
  • Add a brief session note for any significant AI session this week.
  • Check the Rules & Agreements section — are all rules still relevant? Add any new ones.
  • Check Corrections — anything that went wrong this week that isn't logged yet?

Red flag: If your last-updated date is more than 10 days ago, you have drifted. Your AI is working from stale context.

Step 3: Strategy Master Health Check (3 min)

This is a check, not an edit. The strategy master should rarely change.

  • Is the scope still accurate?
  • Are the principles still the right ones?
  • Are the boundaries still correct?
  • Did anything happen this week that challenges the strategy?

If you answered yes to the last question:

  • Is it a minor adjustment? → Update the section, increment version, log in change log.
  • Is it a major shift? → Don't update alone. Think it through explicitly, possibly with AI, before changing the strategy master.

Red flag: If you're updating the strategy master more than once a month, something is structurally unresolved — the strategy isn't stable enough yet.

Step 4: Failure Log Review (2 min)

  • Anything that went wrong this week that should be logged? Add it.
  • Are there any open failures (logged but not resolved)? Advance them.
  • Were there any near-misses — things that almost went wrong but were caught? Log them.

Reminder: Near-misses are as important as failures. If the system caught something before it caused harm, that's worth recording — it shows the system working.

Step 5: Prepare Next Session Context (2 min)

This is the setup for your next AI session.

  • Open questions: Is the list current?
  • Current focus: Does it reflect what you'll work on next?
  • Are there any decisions from this week that your AI doesn't know about yet, that will affect next session's work? Update the running document now.

The test: If you started a new AI session right now and shared just your running document and canonical numbers, would the AI have accurate, current context? If not — update before next session.

Optional: Quick AI Session to Assist the Review

If your project is complex, you can run this review with your AI:

  1. Share your Running Document and Canonical Numbers
  2. Ask: "Help me run the weekly review. Start by checking for any numbers inconsistencies, then help me update the decisions log and open questions."
  3. AI flags inconsistencies; you decide what to update.

This turns the review from a solo ritual into a collaborative audit.

Signs the Review is Working

  • Your AI sessions start faster because context is always current
  • You catch number drift before it reaches deliverables
  • Decisions are traceable — you can always answer "why did we decide that?"
  • Failures are rare because near-misses are caught and fixed

Signs the Review Has Slipped

  • You find yourself explaining things to the AI that you "already told it"
  • Numbers in your work don't match numbers in your files
  • The running document has "last updated" dates more than 2 weeks ago
  • You have open questions in the file that have been resolved but not logged

Run this ritual. Your future self — and your AI — will thank you.