Days 1-4 give you the moves. Day 5 turns them into a weekly rhythm. Under one hour of active Claude time per week. The rest is you doing analyst work — faster, sharper, with less re-explaining.
The four-touch rhythm comes next. But first — the Tuesday block. This is the work. Walk through it.
Claude returns three options, scored, with a recommendation. ~3 minutes.
Claude surfaces something the doc didn't put in headline language — a pricing escalator buried in section 7. ~3 minutes.
Output is a 4-bullet brief in your boss's preferred shape. ~2 minutes.
Output is a paragraph + 3-row table, dollar-anchored. ~2 minutes.
Active Claude time: ~15 minutes. Active you time: ~30 minutes. Output: a synthesized vendor brief, a risk read, two audience-tuned reshapes.
That's what Tuesday looks like once the Vault and the Reads are in place. The work isn't faster because Claude is faster — it's faster because the context, audience tuning, and format don't have to be reconstructed every time.
Most "AI productivity" content tells you to use Claude all day. That's the wrong frame.
The goal is using Claude in specific, scheduled ways so the rest of your week is freed up for the actual analyst work — judgment, stakeholder management, decisions that don't fit in a prompt.
Days 1-4 gave you four moves: synthesize (Decision Frame + Risk Pattern), store context (Status Vault), reshape (audience + format), build Reads (per-stakeholder variants). Day 5 is the schedule that makes them stick.
Decision Frame + Risk Pattern. Stop asking what's in the documents. Start asking what to do about them.
Stakeholders, formats, and reference docs stored where Claude always has them. Stop re-explaining yourself every session.
One source, any audience, any format. Two knobs — audience and format — independently controllable in one short prompt.
Saved prompt variants for each named person you brief. Build once, reuse every week. The asset that compounds the hardest.
A four-touch rhythm. Each touch has one job.
Monday context-update prompt:
On the Friday reset: note anti-patterns — outputs that missed, prompts that didn't work. (Anti-pattern = a prompt that consistently produces bad output, or a stakeholder format that keeps getting pushback. Save the bad version + the fix you used.) This is the difference between a Project that compounds and one that bloats.
Total active Claude time: ~60 minutes per week.
You have two paths. Pick the one that fits your week.
Path A — The full system (if your calendar allows):
Open your calendar. Block four recurring time slots:
Title them whatever fits. The naming doesn't matter. The recurrence does.
Path B — Tuesday only (if your calendar is full):
Skip the calendar blocking. Screenshot the Tuesday walkthrough above and save it as your Monday-morning checklist. Run Tuesday for one week. If it saves you time, add Monday + Thursday + Friday in week 2.
Both paths work. Path A locks in the system. Path B lets the system prove itself before you commit.
A workflow you don't schedule isn't a workflow. It's a hope. But a workflow you adopt all at once is also a hope. Pick the path that fits your bandwidth.
Most analysts using Claude are using it reactively. A doc lands in their inbox, they paste it in, they ask for a summary, they ship the output. That's not a system. That's a faster version of the wrong workflow.
A weekly rhythm turns Claude into infrastructure. You stop thinking about whether to use it. You hit Monday and the Vault is waiting. You hit Tuesday and the Decision Frame is the default move. You hit Thursday and the Reads do the audience work.
That's when the time savings actually compound — when the workflow becomes muscle memory instead of decision.
By the end of this track, you can:
These capabilities take 1-2 weeks of using the system to feel real. Day 5 is the start of the practice, not the finish line.
The Decision Frame won't always return clean options. Sometimes it'll return three, two of which are nearly identical. Sometimes the Risk Pattern will surface a "risk" that's actually noise. The track gives you the prompts. Your judgment decides which signal matters.
This track is a system, not a script. The prompts will get tuned by you. The Reads will get sharper as stakeholders push back. The Vault will compound — or bloat, if you skip Friday.
Treat the rhythm as load-bearing. Skip Monday and Tuesday gets harder. Skip Friday and the Vault rots. The four touches are small because they're meant to be sustainable, not impressive.
Give it two weeks of real use. The system gets sharper as you tune it — your first week will feel slower than your fourth.
That's the analyst version of using Claude. Not more. Sharper.
The Status Vault, the Reshape Prompt, and the four-touch weekly rhythm — packaged as a one-page reference you can keep open while you work.
Get the Work OS →