The Weekly Operating Rhythm

Three nested cadences, daily, weekly, monthly, that keep the system from rotting.

Foundry is not a tool you use whenever. It's an operating system, and operating systems require rhythms. Without rhythm, every feature degrades, captures pile up, decisions accumulate without verdicts, intentions drift into vague. With rhythm, the same features compound.

The cadence is three nested loops:

• Daily (2 min). Today scan. Orient.


• Weekly (15 min). Review. Close last week's loops, place next week's bets.


• Monthly (30 min). Strategic check-in. Pattern-spot across the four diagnostic surfaces.

Same five-step pattern, every working morning, in the order they appear on Today:

1. Focus project, confirm you still believe in it. (10 sec)
2. Overdue revisits, record verdicts or reschedule. Do not just dismiss. (~30 sec each)
3. Due commitments, close anything done; flag anything that will slip. (~1 min)
4. Weekly intentions, check anything completed. The check is the small dopamine hit. (15 sec)
5. Stale captures, if 3+, schedule a triage block this week. Don't process them now. (15 sec)

Under two minutes. The point is orientation, not work. The actual work happens on the relevant detail pages.

If the daily scan consistently runs over 5 minutes, the system is overloaded somewhere upstream, too many active commitments, decisions piling up unresolved, stale captures growing. Fix the upstream condition, don't extend the time block.

Pick a slot (Friday afternoon or Sunday evening are the two that work). Make it a 30-minute calendar block even though the review itself takes 15, the buffer is for reflection that doesn't fit in the form.

The five-step ritual (covered in detail in the Weekly Review module): Commitments → Reflect → Clear Queue → Wins & Learnings → Intentions.

The two non-negotiable lessons:

Promote one learning to a LESSON_LEARNED asset. Every week. This single habit produces ~50 durable, searchable lessons per year. Skipping it for "speed" saves 90 seconds and costs the entire compounding effect.

Write intentions as hypotheses, not tasks. "Finish the landing page" is task-shaped, you check it off when done. "Get the landing page to a state I'd show 10 people" is intention-shaped, you check it off when your judgment says you're satisfied. Intentions teach you about your judgment. Tasks just measure throughput.

Block 30 minutes on the last Friday of the month (or first Friday of the next). The sequence:

1. Overview (5 min), read every number. Name the single weakest signal.
2. Focus (5 min), confirm Focus Lane still belongs at Layer 2; audit energy mix.
3. Recommendations (10 min), resolve every High alert. Note any that have fired multiple months in a row.
4. Review History (10 min), read the last 4 weekly snapshots in sequence. Name one pattern you didn't see in the moment.

End by logging a Decision titled "[Month] check-in" with a one-paragraph summary of what you found and what you'll change. The decision is the artifact that makes the check-in stick. Without it, you've just looked at numbers and forgotten them by Tuesday.

If you have Pro, run AI Pattern Detection quarterly during the check-in. It surfaces patterns across 60 days of data that no human reads through manually.

Weekly break (skipped 1 review): Recoverable. Do this week's review now; don't try to retroactively reconstruct last week's. Note the skip in the reflection field.

Weekly break (skipped 2+ reviews): Dangerous. By week three, the cadence is gone and reinstating requires deliberate effort. The fix: skip the perfectionism, do a partial review today, even if it's just intentions + commitments. The act of doing breaks the spiral.

Monthly break (skipped 1 month): Almost no consequence in the moment. The cost shows up two months later when you notice structural drift you would have caught.

All three breaking simultaneously: This is the start of every "I stopped using Foundry" story. Reinstate the daily scan first, it's the lowest-cost re-entry, and the others follow.