| name | mind-time-review |
| description | Run structured weekly, monthly, or quarterly reviews of digital ATTENTION and information diet, using Rize automatic time tracking. Measures the attention leak (social media, entertainment/streaming, doomscroll news), focus quality and fragmentation, slow long-form vs short-form dopamine consumption, and desktop screen-time load, scored against the Lifestyle screen-time target, the strategy principles ("Distracted, not focused -> deep work / mono-tasking", "Short videos -> consume slow and long-form", "Spend more time offline"), and the Mind information-diet references (Digital Minimalism, dopamine culture). Trigger on attention review, focus review, screen-time review, digital wellbeing / digital minimalism check, distraction audit, information-diet review, or "how healthy was my attention this week/month". For the work-output lens (Zone of Genius, meeting load, deep-work productivity) use `work-cadence-review` instead. |
| compatibility | [{"tool":"mcp__rize__get_current_user"},{"tool":"mcp__rize__list_my_apps_used"},{"tool":"mcp__rize__get_my_time_allocation"},{"tool":"mcp__rize__list_my_time_entries"}] |
Mind - Time Review (Attention)
Primary entrypoint for reviewing the health of digital attention and information diet, powered by Rize.
This is the attention-hygiene lens on time; the work-output lens lives in work:work-cadence-review. They read the SAME Rize data and are cheap to run together — offer both when the user asks for "a time review" unqualified.
Review Types
- weekly review (this ISO week vs last ISO week)
- monthly review (this calendar month vs last)
- quarterly review (multi-month trend; earliest month = baseline)
Use ISO week boundaries for weekly and calendar months for monthly. No rolling "last 7 days".
Reasoning Order
- Read
000 OS/2 Simple strategy 2026 — the attention-relevant lines: "Distracted, not focused -> Deep work sessions with mono-tasking", "Short videos -> Consume slow and long-form content", and core principle "Spend more time offline and outside in places that give me energy".
- Check the screen-time targets in
300 Areas/4 Lifestyle/0 - Red - Hobby & Fun — read the current daily mobile screen-time and phone-use targets from that file rather than assuming values.
- Read
300 Areas/3 Mind/2 - Orange - Goals & Performance (the information-diet / LMS hub) and its linked references [[Digital_Minimalism-_Choosing_a_Focus]] and [[The rise of dopamine culture]].
- Only then interpret the Rize data.
Data Sources
rize MCP (mcp__rize__*) — automatic desktop time tracking. Source of truth for what happened on the laptop.
- If a phone/mobile screen-time number exists (Screen Time export, manual estimate), use it — Rize cannot supply it.
Data Reality Rules — READ BEFORE EVERY REVIEW
- Build from
list_my_apps_used, not from labels. Label coverage is sparse-to-absent historically (Mar-May 2026 were 100% "Unassigned"; Jun ~67% unassigned). Rebuild categories yourself.
- Do NOT report Rize's
is_focus% as your focus number. Rize marks Slack, LinkedIn, and x.com as is_focus: true. That inflates "focus". Compute your own focus vs leak split from the Bucket Taxonomy and report the gap between the two — it is the review's honesty anchor.
list_my_events is flaky (7-day cap, frequent errors). Use entry_count per tracked day as the fragmentation proxy.
- Rize is desktop-only. It cannot measure phone doomscrolling. The screen-time target is a phone target — state plainly that Rize does not cover it, and treat the desktop leak as a lower bound on total distraction.
list_my_time_entries narrates what the sessions were; sample it to distinguish, e.g., long-form learning from short-form scrolling within the same "browsing" time.
Bucket Taxonomy
From list_my_apps_used (app-name overrides win over category):
| Bucket | Rule | Attention read |
|---|
| Focused work | Code, AI tools, Writing/docs, Design, Research/learning | The good — sustained attention |
| Meetings | Video Conferencing | Structured but not solo-focus |
| Comms | Communications, Email, Messaging | Fragmentation driver |
| Social media | Social Media, Linkedin; x.com | Leak — dopamine / infinite scroll |
| Entertainment | Entertainment (netflix, youtube, streaming) | Split long-form (Netflix film) vs short-form (YouTube/shorts) per "consume slow, long-form" |
| News | News (news sites, current-affairs media) | Leak when doomscrolling; note news-cycle pull |
| Admin/personal | Scheduling, Hiring, shopping, personal sites | Neutral logistics |
| Browser (unattributed) | bare Brave/Chrome | Neutral |
Attention leak = Social media + Entertainment + News. Report it in hours and as % of tracked desktop time.
Metrics To Compute (per period, with delta)
- Total tracked desktop hours; average/day. (State: phone not included.)
- Attention leak (social + entertainment + news) in hours and % of tracked — the headline.
- Social-media hours, broken to platforms (threads, x, linkedin, instagram, facebook).
- Entertainment hours, split long-form vs short-form (Netflix/long film vs YouTube/shorts/streaming binges) — score against "consume slow and long-form".
- News/doomscroll hours.
- Focus ratio — focused-work hours / tracked hours, YOUR rebuild, next to Rize's inflated
is_focus%.
- Fragmentation — entry_count / tracked-day; comms hours as a switching driver.
- Longest-standing distraction — the single biggest leak app and its trend across periods.
Workflow
- Set the review + comparison windows.
- Read the strategy attention-lines, the Lifestyle screen-time target, and the Mind information-diet references before concluding.
- Parallel Rize pull:
get_current_user, list_my_apps_used (review + comparison windows), get_my_time_allocation (entry_count/tracked totals), and a list_my_time_entries sample.
- Rebuild buckets; compute the leak, focus ratio, long-vs-short-form split, and fragmentation for each window.
- Compare current vs prior equivalent period.
- Score against the screen-time target, the "mono-tasking / slow long-form / offline" principles, and the digital-minimalism references.
- Back-reference the prior review's priorities; grade follow-through.
- Produce the structured review; save the file(s).
Recommendation Rules
- One to three high-leverage priorities. Prefer subtractions (block a site, cap a platform, batch comms) over vague "be more mindful".
- Tie each to a specific leak in the data and a specific principle or target.
- Separate observation / interpretation / assumption. Vault voice: sentences < 25 words; numbers over generalities; banned words — delve, leverage, robust, navigate, journey, ecosystem.
Output Contract
- review scope + comparison windows (+ "Source: Rize desktop only; phone attention not captured — leak figures are a lower bound").
- a scorecard delta table (leak buckets + focus ratio across periods, Trend column for multi-month).
- for multi-month trends, open with a narrative headline reframing the numbers.
- attention-leak, social-media, entertainment (long vs short form), news, and fragmentation sections.
- the rebuilt-focus% vs Rize-is_focus% reconciliation.
- alignment with the screen-time target, attention principles, and information-diet references.
- top 3 priorities; risks/caveats (desktop-only lower bound, label sparsity); numbered next actions.
Numbers first, then interpretation, then recommendation.
File Saving
Weekly and monthly reviews always create report files. Do not ask whether to save.
mkdir -p "$PERIODICS_ROOT/Monthly/$MONTH_NAME"
mkdir -p "$PERIODICS_ROOT/Weekly/Week $ISO_WEEK"
Example paths:
- monthly:
$PERIODICS_ROOT/Monthly/06 June/2026-06-mind-time-review.md
- weekly:
$PERIODICS_ROOT/Weekly/Week 11/2026-week-11-mind-time-review.md
H1 + bold metadata block (Review period / Comparison period / Source / Generated / Related). No YAML frontmatter on review outputs.
Resources
Pull the Mind information-diet references from 400 Resources/ (Digital Minimalism, dopamine culture) when a recommendation needs backing.
Tone
Honest attention audit. Numbers first, no moralizing, end with concrete subtractions.
Schema
Reference ../../schemas/time-review.json. Set lens: "mind-attention".