Stop assigning work. Start receiving proposals.
This is the destination unit of the seven-phase journey. The relationship between you and Claude finally inverts — you stop opening the conversation and start being briefed.
For six units, every interaction has started with you. You typed the prompt. You triggered the Skill. You ran the workflow. Phase 7 changes that for the first time: Claude runs on a schedule, watches what you said to watch, and surfaces things to you when they need your attention.
The unit isn't aspirational fluff. It's a real design problem. What should Claude watch? What deserves your attention and what doesn't? How does Claude propose without overreaching? How do you stay the decision-maker when Claude is doing more of the noticing?
What you'll walk out with
A clear understanding of what makes a great Phase 7 setup — and what makes a bad one. A four-move pattern for partner-quality Skills (W-N-P-W). Two worked examples of professionals who genuinely live in Phase 7. And two hands-on builds: your first "noticer" Skill and a proposal format you can use across every Phase 7 routine you ever build.
What Unit 7 covers
- The Phase 6 ceiling — why even a great workflow setup still leaves you initiating.
- The mindset shift — from asking to being asked.
- W-N-P-W — Watch, Notice, Propose, Wait. The four moves of a partner Skill.
- Scheduled and triggered Claude — when partnership becomes recurring.
- The noticer — what to watch, what to ignore, how to calibrate.
- The proposal format — how Claude packages "here's what's happening, here's what I'd do, what do you want?"
- Two worked examples — Rachel's unprompted Tuesday email; Marcus's Monday morning briefing.
- Three Phase 7 mistakes — over-watching, proposing without action, over-trusting proposals.
- Two hands-on builds — design your noticer; design your proposal format.
- The frontier — what's beyond Partner, and the hook into Unit 8.
Every chain still starts with you.
Phase 6 made your workflows end-to-end. Phase 6 didn't make them self-starting. You're still the one opening Claude. That's a smaller ceiling than the previous five — but it's real, and it's the last one.
Watch yourself for a week running Phase 6 workflows. The pattern is consistent: you open Claude, you trigger the routine, you review the result, you close. The routine is autonomous. The starting it isn't.
For most working professionals, this works perfectly well. Phase 6 is fast, safe, and reclaims hours per week. You can comfortably stop here — many people will. But there's a kind of leverage that only shows up in Phase 7: the things you didn't think to ask about.
Three signs you're hitting the Phase 6 ceiling
1. You discover problems after the fact.
The customer who churned was at risk for three weeks. You'd have caught it if you'd looked. You didn't look — because you didn't have a workflow scheduled to look. Phase 7 is what watches the things you'd otherwise miss.
2. Your "I should check on X" list is growing.
You mean to check on the funnel weekly. You mean to look at the customer health board on Tuesdays. You mean to glance at the cash position before payroll. These are the to-dos that fall off — exactly the kind of low-frequency, high-importance watch points Phase 7 is built for.
3. You wish Claude would tell you something instead of waiting.
You've stopped wondering what Claude can do. You've started wondering why you have to keep being the one to ask. That wish is the Phase 7 doorway.
The honest version
Phase 7 only makes sense once you have solid Phase 6 workflows running. Without them, "scheduled Claude" is just noise. With them, Phase 7 is the layer that picks up the things your workflows didn't ask about.
From asking to being asked.
The shift sounds dramatic. In practice it's small and specific. You go from typing "what's going on with X?" to receiving an email that says "X is happening. Here's what I'd do. What do you want?"
For the first six phases, your judgment activated on demand — when you decided to deploy it. In Phase 7, your judgment activates when Claude surfaces something. Same judgment. Different trigger.
The slogan that does the work
Watch. Notice. Propose. Wait.
The four moves of a partner Skill. The Watch happens on a schedule. The Notice is what Claude flags. The Propose is what Claude recommends. The Wait — the part most professionals get wrong — is where Claude stops and lets you decide.
The "Wait" is non-negotiable
In Phase 7, Claude could conceivably propose and act. Most early adopters who skip the Wait step regret it within a month. The customer outreach that went out without you approving it. The calendar invite that landed on the wrong person's calendar. The Slack post that went up overnight and misread the room.
The Wait is what preserves you as the decision-maker. Claude does the watching, the noticing, the proposing. You do the deciding. That's partnership. Without the Wait, it's something less valuable and more dangerous: delegation without supervision.
What changes in your week
Marcus's Tuesday morning, pre-Phase 7: he opens Claude, asks for the renewal-risk update, reviews it, closes Claude. About 12 minutes.
Post-Phase 7: he wakes up to an email from Claude. "Three customers crossed into Medium renewal risk this week. Here are the drivers. Here's what I'd do for each — and one I think needs a different conversation than the template. Want to handle any of these today?"
Same work. Same judgment. Different entry point. He's not opening Claude. Claude is opening to him.
W-N-P-W — the four moves of a partner Skill.
Every Phase 7 setup has four moves. They differ from R-R-C-A in one fundamental way: the first move is Watch on a schedule, not Read on demand. The trigger isn't you. It's the calendar.
Watch
Claude wakes up on a schedule — daily, weekly, or in response to a triggering event — and pulls from your connected systems. The same Read step from Phase 5, except you're not triggering it. The clock is.
The schedule matters. Daily is too much for most signals; weekly is the sweet spot for most working professionals.
Notice
The data is loaded. Now Claude filters: which of these signals actually deserves your attention? The answer comes from rules you set — your noticer logic. "Customers crossing into Medium risk." "Pipeline movement >10%." "Any project trending past deadline."
A great noticer is calibrated. A bad noticer surfaces everything and gets ignored.
Propose
For each thing worth surfacing, Claude doesn't just describe it — it proposes a response. The draft of the outreach email. The recommended decision and the reasoning. The "I'd do X, but you might prefer Y for these reasons."
Description-only is a "FYI." A proposal is a partnership move.
Wait
Claude packages the noticings and proposals into a single brief — email, Slack DM, dashboard, whatever format you specified — and stops. Waits for your reply. You approve, you redirect, you handle it yourself. Then Claude resumes its watch.
The Wait is the partnership contract. Without it, you've delegated something you shouldn't have.
How W-N-P-W differs from R-R-C-A
The patterns share the same DNA — Claude reads, reasons, then either confirms or proposes, then acts or waits. The crucial difference: in Phase 5, you initiated the cycle by triggering the Skill. In Phase 7, the schedule does. Your judgment shifts from "I'd like Claude to do X right now" to "Claude noticed Y; here's my decision."
The Phase 7 contract
You are still the decision-maker. Claude does more of the noticing. That sentence is the entire Phase 7 governance model. Lose the second half and Phase 7 doesn't pay off. Lose the first half and it costs you.
Rachel's Tuesday morning email — the one she didn't ask for.
Rachel is executive director of a regional nonprofit. She has Phase 5 Skills and Phase 6 workflows handling grant reporting, board prep, and weekly fundraising recaps. The thing she kept missing: the slow drift of donor relationships.
She built a Phase 7 noticer specifically for donor engagement. Here's the full setup:
Every Tuesday at 6am, Claude pulls:
- Top 25 donors' engagement: opened any of last 4 newsletters? clicked anything? attended any event in the last quarter?
- Public news, IRS filings, and major social posts from each donor's foundation or business
- Any direct correspondence between Rachel and the donor in the last 60 days
- Our program activity: where their funding went, what outcomes were achieved
Surface any donor where:
- Engagement dropped below their normal baseline (no clicks in 4+ newsletters when they normally click)
- A public signal suggests strategic shift (new program focus in their 990, executive changes, new grants outside our space)
- It's been >90 days since Rachel last had direct contact
- A program outcome is publishable and notable (positive story that matches their funding interest)
For each surfaced donor, draft:
- A one-paragraph read of what's happening with them
- The recommended response — a personal email, a thank-you note, an invitation to a site visit, an outcome story to share, or "no action; just wanted you to see this"
- If outreach is recommended, a draft of the email in Rachel's voice
Package the surfaced donors (typically 0–5 per week) into a single email to Rachel by 7am Tuesday with subject line "Donor watch — week of [date]." Each donor gets a short section: the read, the recommendation, the draft. Rachel responds with send, edit, skip, or handle differently for each item.
What this Skill is actually doing for Rachel
The first month she ran it, two donors surfaced that she would not have flagged on her own. One was drifting — Claude caught the engagement drop. Rachel sent a personal note about a recent program outcome. The donor wrote back, re-engaged, made a larger pledge the following quarter.
The second was the opposite — a donor whose 990 showed they were shifting toward youth programs. Claude proposed drafting a specific outreach about Rachel's youth program work. Rachel sent it. New funding conversation opened.
Neither outreach would have happened without the watch. Rachel didn't have time to manually track engagement signals across 25 donors weekly. Claude did. The judgment stayed Rachel's. The noticing didn't.
Marcus's Monday morning briefing.
Marcus already had Phase 6 workflows running his pipeline recap. He realized he was still spending the first hour of Monday catching up on what happened over the weekend, what's at risk this week, what's worth attention. Phase 7 took it.
His partner Skill runs Sunday night. Marcus wakes up Monday to a state-of-the-business briefing already in his inbox:
Sunday evening, Claude pulls:
- Salesforce: full pipeline state, every deal's current stage, last-touched date, and any change since last Sunday
- Sales activity: rep-level outbound activity, response rates, meetings booked
- Customer success: any account moved to At-Risk during the week, any expansions discussed
- Calendar: Monday's meetings — Marcus's and his AEs' — and what deals they touch
- Slack: any flagged channel messages tagged "deal alert" or escalated
Filter to:
- Deals that went silent >7 days that previously had weekly activity
- Reps whose outbound dropped >30% from their 4-week baseline
- Any At-Risk movement that wasn't surfaced during the week
- Any Monday meeting on a deal in late stage where the deal hasn't been touched in 5+ days
- Any positive surprise (large new opportunity, unexpected expansion)
For each surfaced item:
- One-line read: what's happening
- The recommended Monday action: who to talk to, what to ask, what decision to make
- For rep performance items, the proposed coaching angle
- For meeting prep items, a 3-bullet brief on the deal status
Compose a single Monday briefing email sent by 6:00am, structured: Headline · Things needing my attention today · This week's watch list · Positive notes. Marcus replies with assignments — "handle this myself," "draft this for me," "raise this in the standup," "ignore this one and tell me why next week."
Why this works for Marcus
Before Phase 7: 45–60 minutes Monday morning catching up. He missed things weekly. He sometimes triggered the standard pipeline recap Skill and read it. Sometimes he didn't.
After Phase 7: he wakes up to 4–6 items already filtered, already proposed. He spends 15 minutes responding to each line. By 7am the standup is prepped, the coaching conversations are scheduled, the two at-risk deals have a plan. Marcus stopped being his own analyst. He became the decision-maker the analyst is briefing.
Reflection · 3 minutes
If you had a Phase 7 briefing waiting for you Monday morning, what THREE things would you most want it to surface? Be specific — these are your first noticer rules.
Three mistakes that turn partners into noise.
Phase 7 is the easiest phase to mis-design. The mistakes all look like rigor or thoughtfulness in the moment. The cost shows up over weeks.
Mistake 1 — Over-watching.
You set up Phase 7 and you want to make sure you don't miss anything. So you set the noticer to flag generously. Within two weeks, the Monday briefing has 30 items, half of them low-signal. You start skimming. By week four, you stop reading. Phase 7 quietly became noise.
Mistake 2 — Proposing without action.
The noticer surfaces well. The brief describes what's happening clearly. But it doesn't propose anything specific — "FYI, customer X is showing signs of churn." You read it, agree it's important, do nothing, and the customer still churns. The noticing was right. The proposal was missing.
Mistake 3 — Over-trusting the proposals.
The first month is great. Claude's noticings are sharp. The proposals are mostly right. You approve them quickly. Then one week you approve a proposal that was wrong — the customer wasn't churning, the rep wasn't underperforming, the situation needed a completely different read. You weren't actually reviewing anymore.
The Phase 7 debugging move
When the briefing stops being useful, walk through W-N-P-W. Watching too much? Notice criteria too loose? Proposals too generic? Are you actually Waiting — or are you rubber-stamping? Phase 7 is the most maintenance-sensitive phase. Plan to recalibrate quarterly.
Your turn.
Two hands-on builds. First a noticer — the rules that tell Claude what to surface. Then a proposal format — the structure Claude uses for every Phase 7 briefing.
Design your noticer.
Pick one area of your work where you'd benefit from a weekly briefing — customer health, sales pipeline, team performance, project status. List the signals Claude should watch for. Be precise — the difference between "anything important" and "engagement drops below baseline" is the difference between noise and signal.
One way to design a "team health" noticer
Watching what: 1:1 calendar invites and notes (frequency, who's missing); team Slack channels (volume changes per person, response times); project tools (assignments per person, blocker frequency); HR data (PTO patterns, recent reviews).
3 signals to surface: (1) Any team member whose 1:1 has been missed or rescheduled twice in a row. (2) Any team member whose Slack activity dropped >40% from their 4-week baseline. (3) Any team member assigned >1.5x their usual workload this sprint.
3 things to ignore: (1) Activity changes that correlate with logged PTO. (2) Slack drops on days the team had offsite or no-meeting days. (3) Workload spikes during the first/last week of any quarter.
Cycle: Weekly, Monday at 6am. One briefing email with surfaced people only — no email if nothing surfaces.
Notice what's not in there. "Be helpful." "Watch for anything important." Both are noise generators. The model answer is precise — specific signals, specific thresholds, specific exclusions. Precision is the difference between a Phase 7 briefing you read every week and one you start to ignore.
Design your proposal format.
Across every Phase 7 routine you ever build, you want the briefings to feel the same. Same shape. Same length. Same structure. So your eye knows where to look. The format is the contract.
Your standard proposal format.
Walk through how you'd want every Phase 7 briefing structured. The sections. The length per section. The expected response actions. What "easy to act on" looks like in your inbox.
One way to design your proposal format
Headline: One sentence at the top: "[N] items need your attention this week." If N = 0, no email — just a confirmation Slack DM that the watch ran clean.
Per item structure: Three short blocks per surfaced item. (1) "The read" — 2 sentences describing what's happening and why it matters. (2) "What I'd do" — the proposed action in plain English, including the draft if the action involves communication. (3) "What I need from you" — the specific reply needed.
Response actions: For each item, I reply with one of: send (proceed with the draft), edit: [my changes], skip (don't act this week), I'll handle this (I'll do it personally, don't draft), or tell me more about [aspect]. Claude processes my replies and either acts, skips, or comes back with more detail.
Delivery: Email to my work address, sent by 6am the morning I'm watching. Subject line follows pattern: "[Watch name] · Week of [date] · [N] items." No emails if nothing surfaces.
Format matters as much as content. A great Phase 7 briefing makes the decision easy — eye scans the headline, eye sees the count, eye reads each item in 30 seconds, reply in plain English. If you have to think hard about what to do, the format is wrong.
Past Partner — what's next.
Phase 7 was supposed to be the destination. For most working professionals, it is. For the small group who get here and want to keep building, there's a frontier — and Unit 8 is your map of it.
By the time you're running well in Phase 7, you've crossed from "user of Claude" to "designer of an AI partnership." You have Custom Instructions. Projects. Skills. Operator Skills with R-R-C-A. Workflows with checkpoints and escalation. Partner Skills with W-N-P-W. The seven phases of the framework all run for you.
What's left? Things the framework didn't name — not because they're hidden, but because they only become useful once you've mastered everything before them.
The frontier is real, and small
Unit 8 — "Beyond Partner" — covers the advanced patterns that the next generation of working professionals will use. The Knowledge Compound (Claude as a thinking partner that builds institutional memory in real time). Multi-Claude workflows (multiple Claude instances coordinating on complex deliverables). The Personal MCP Server (custom integrations to systems no off-the-shelf connector handles). The Council (Claude in adversarial roles for stress-testing decisions). The Decision Atlas (Claude as your decision archive and pattern-finder).
These aren't theoretical. Working professionals are using them today. They are also where the leverage of AI starts compounding instead of just scaling.
Who Unit 8 is for
If you're new to Claude, Unit 8 will read as science fiction. If you finished Units 1 through 7 and at least one of your Phase 7 routines is running on autopilot — Unit 8 is a working playbook. It is the unit that turns the seven-phase framework into a personal AI infrastructure.
Reflection · 2 minutes
If Phase 7 is where you spend your time within a year, what would the next move look like for you? What's the work AI could be doing that the off-the-shelf framework hasn't named?
Are you ready for the frontier?
Three questions to diagnose whether Unit 8 will land as a working playbook or as future reading.
1. How many Phase 7 routines are you running today?
2. How would you describe your current relationship with Claude?
3. How much of your week's work is now scheduled or surfaced by Claude rather than initiated by you?
Your readiness
Answer all three questions above to see your readiness for the frontier (Unit 8).
Seven questions.
One per major idea in this unit.
0 of 7 answered · 0 correct so far
1. The Phase 6 ceiling — the limit even great workflows can't push through — is:
2. The four moves of a partner Skill are:
3. What is the role of the "Wait" step in a partner Skill?
4. What separates a great noticer from a noisy one?
5. The most common Phase 7 mistake is:
6. The role of the proposal format is:
7. Beyond Phase 7, Unit 8 covers patterns like:
Finish the knowledge check to see your score.
Your noticer design and proposal format are saved.
The new thought process.
Stop assigning. Start receiving.
The relationship inverts. Claude opens to you. You decide.
W-N-P-W.
Watch on a schedule. Notice with precision. Propose with action. Wait for your call.
The Wait is non-negotiable.
You stay the decision-maker. Claude does more of the noticing. That's the partnership contract.
Precision > volume.
A great noticer surfaces few things, right things. A noisy noticer gets ignored.