felipefontoura/axi25-core/axi25-plan

Weekly Plan

Forward-looking weekly planning. While axi25-review asks "what happened?", this asks "what will I make happen?"

Voice: Respond in the language the user is speaking (see 90-system/references/user-profile.md).

When to Use

• Start of a new week (Sunday night or Monday morning)
• After a review (review feeds data, plan feeds action)
• When the user explicitly asks to plan the week
• When the user says "plan W##" or similar

Do NOT trigger on: "how was my week", "summary", "review" — those route to axi25-review.

Relationship to axi25-review

Ideal flow: review Sunday → plan Monday (or same session: review first, then plan). The plan references the review's conclusions.

Steps

1. Gather context (READ)

Read in order:

1. Last weekly review: 40-journal/weekly/YYYY-W(N-1).md — conclusions, open loops, "next week focus"
2. Last weekly plan (if exists): 40-journal/weekly/YYYY-WNN.md — planned vs. what the review found
3. Active area pages: 20-wiki/areas/*.md — current goals and signals
4. Active projects: 30-projects/active/*.md — status and next steps
5. Active decisions: 20-wiki/decisions/*.md (status=active) — especially decision-routines with review dates
6. Patterns: 20-wiki/patterns/*.md — which patterns are live and need accounting
7. Recent journal entries (last 3-4 days): 40-journal/daily/ — mood, energy, momentum
8. Recent study notes (last 3-4 days): 25-studies/ — study in formation, working theories, candidate promotions

2. Dialogue with the human (ASK)

This is the key difference from review. The plan requires the human's intent. Present a draft and ask for confirmation/correction on:

• Priority ranking: "Of these areas/projects, which are top priority this week?"
• Capacity estimate: "Realistically, how much energy/time do you have?"
• Non-negotiables: "Any hard commitments this week? (appointments, deadlines, family)"
• Explicit NO list: "What are you deliberately NOT doing this week?"
• Study focus: "Which ideas need to mature this week without becoming projects yet?"

If the human already provided direction in their prompt (e.g., "focus on the launch this week"), use that as the primary signal and confirm the rest.

3. Generate the plan (WRITE)

Write to 40-journal/weekly/YYYY-WNN.md — same file where the review lives, but with type: weekly-plan (or weekly-plan-review if both exist).

---
type: weekly-plan
week: YYYY-WNN
period: YYYY-MM-DD to YYYY-MM-DD
based_on: YYYY-W(N-1)
---
 
# Weekly Plan — WNN (YYYY-MM-DD → YYYY-MM-DD)
 
## Thesis of the week
 
1-2 sentences: what is the core hypothesis or intention this week?
e.g., "Execution week: ship the launch and read the market's response."
 
## Goals by area
 
### Work
-[ ] Goal 1 (measurable)
-[ ] Goal 2
 
### Health
-[ ] Goal (e.g., sleep ≤23:30 on 5/7 days)
 
### Family
-[ ] Goal
 
### Finance
-[ ] Goal
 
### Learning
-[ ] Goal (optional — leave empty if not a focus week)
 
### Spirituality / Philosophy
-[ ] Goal or discernment focus (optional)
 
## Day-by-day distribution
 
| Day | Focus | Key action |
|---|---|---|
| **MON DD** | AREA | Key action |
| **TUE DD** | AREA | Key action |
| ... | ... | ... |
| **SUN DD** | REVIEW | Review WNN + plan W(N+1) |
 
## Non-negotiables
 
-**Sleep ≤HH:MM** — target N/7 days
-**Morning HH-HH = deep-work block** — no phone, no email
-**Fixed commitment** — e.g., a weekly appointment
-**Daily journal** — any format
 
## What NOT to do this week
 
-❌ Item 1 — reason
-❌ Item 2 — reason
-❌ Item 3 — reason
 
Be explicit and aggressive. Every "not doing" frees energy for what matters.
 
## Success indicators
 
(How will the review know the week worked?)
 
1.Measurable indicator 1 (e.g., "launch shipped by MON?")
2.Measurable indicator 2
3.Measurable indicator 3
4.Qualitative indicator (e.g., "did the routine feed or kill creativity?")
5.**Thesis validated or refuted?** — is the week's thesis true?
 
## Decision-routine (if any)
 
If a `decision-routine-W(N-1)` is active with a review date in this week:
 
-List indicators to verify
-Note: "Review of decision-routine scheduled for YYYY-MM-DD"
 
## Connected
 
-[[relevant-project]]
-[[relevant-decision]]
-[[relevant-pattern]]
-[[previous-week-review]]

4. Update related pages

• Area pages (20-wiki/areas/*.md): add "Weekly plan WNN: ..." to Recent Signals if the plan defines a new intention
• Active decisions (20-wiki/decisions/*.md): if a decision-routine hits its review date this week, add a note
• Project pages (30-projects/active/*.md): update Status with the week's intended milestone
• `index.md`: add/update the weekly plan entry
• `log.md`: append entry

5. Confirm with the human

Present the key structure — thesis, top 2-3 priorities, what's explicitly NOT happening, non-negotiables — and ask: "Anything to adjust before we lock it in?"

Anti-Patterns

• Don't plan without reviewing last week — the plan must acknowledge what happened or didn't
• Don't create wishlists — every goal needs a day assigned or a measurable outcome
• Don't skip the "NOT doing" list — this is where the real prioritization happens
• Don't plan more than 3 focus areas per week — dispersion kills execution
• Don't ignore patterns from last week — if sleep was broken, the plan must account for it
• Don't overwrite a review — if a review for the same week exists, merge plan sections into it rather than replacing

Notes on File Conventions

• Weekly plans live in 40-journal/weekly/YYYY-WNN.md — same location as reviews
• If both review and plan exist for the same week, they coexist in the same file (review first, plan after, separated by an HR ---)
• The type frontmatter reflects the primary purpose: weekly-review, weekly-plan, or weekly-plan-review