agentleFS
eugenezhangco-speceugenezhangco-spec/athena/.claude/skills/weekly-review/SKILL.md

weekly-review

Skill0 starsChanged 7 months agoMITView on GitHub

Runs a structured weekly review session where the assistant asks the user about their week, updates all context files, audits backlogs, and sets priorities for next week. Use this skill whenever the user says 'let's do the weekly review', 'weekly review', 'weekly check-in', 'let's review the week', 'end of week review', or 'how did this week go'. Also activates when the user wants to audit and clean up context files, update stale metrics, or align on priorities. Do NOT use for daily planning (use day-planner). This is the assistant and user sitting down once a week to get everything current.

99 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use quarterly-review, daily-plan and 8 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • Phoenixrr2113Skill

    quarterly-review

    Quarterly review — strategic look at the last 3 months. Career trajectory, project health, life balance, big decisions ahead. Runs 1st of Jan/Apr/Jul/Oct.

    Also in 3 projects that use this

  • steptianSkill

    daily-plan

    每日待办计划与日程管理。触发词:'今天要做'、'今日计划'、'今日待办'、'今天安排'、'安排一下'、'加个任务'、'记个任务'、'今日复盘'、'今日回顾'、'今天总结'、'本周计划'、'周总结'、'开始工作'、'上班了'、'完成了'、'加个日程'、'记个日程'、'改一下时间'、'日程调整'、'面试'、'开会'、'约了'。支持四象限优先级、时段分配、固定日程管理、智能时间估算、每日/每周复盘。当用户提到有明确时间的安排(如'下午两点面试'、'10点开会'),自动识别为日程并写入日程表。

    Also in 3 projects that use this

  • seedprodSkill

    daily-brief

    Send a morning briefing via Telegram with weather, news, and a motivational quote. Designed to run via launchd for automated daily delivery.

    Also in 5 projects that use this

  • claude-office-skillsSkill

    meeting-notes

    >

    Also in 8 projects that use this

  • Phoenixrr2113Skill

    morning-brief

    Full morning brief — calendar, email, meeting prep, execute actionable items, create follow-up tasks. Runs at 8:03 AM or when Randy asks for a morning update.

    Also in 4 projects that use this

  • JerovinoSkill

    process-inbox

    Processes tasks.md inbox items by triaging each one to its proper location in the vault

    Also in 3 projects that use this

The file

Show its rules and text

We split this skill into the rules it gives an agent

24 rules. 0 are common, 24 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 85 edited versions we read keep these.

  • 24

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read these files before starting:

    line 19
  2. Distinctive:

    Run every phase in order. Each phase has specific questions the assistant asks the user.

    line 34
  3. Distinctive:

    Update snapshot.md flags with whatever the user reports.

    line 58
  4. Distinctive:

    Do this for each task category:

    line 64
  5. Distinctive:

    Read this week's day-ledger.md entries first.

    line 88
  6. Distinctive:

    If project hours are under target, name it: "You got [X] project hours this week.

    line 92
  7. Distinctive:

    If work expanded beyond normal, name it: "Work took [X] hours this week.

    line 93
  8. Distinctive:

    If an observation has been confirmed twice, promote it to patterns.md.

    line 102
  9. Distinctive:

    If a pattern is no longer true, update or remove it.

    line 103
  10. Distinctive:

    Ask: "Anything you have noticed about yourself this week that I should know?"

    line 104
  11. Distinctive:

    Check inbox.md: any pending items that should be actioned or archived?

    line 110
  12. Distinctive:

    Check snapshot.md: are all flags still relevant?

    line 111
  13. Distinctive:

    Check if any on-demand files were loaded in 5+ consecutive sessions (should they be promoted to always-loaded?)

    line 112
  14. Distinctive:

    Check if any always-loaded content was not used in recent sessions (should it move to on-demand?)

    line 113
  15. Distinctive:

    Ask: "Anything about how I work that felt off this week?

    line 115
  16. Distinctive:

    After the user answers:

    line 127
  17. Distinctive:

    Update snapshot.md with new flags and next-week priorities.

    line 128
  18. Distinctive:

    Update patterns.md if new observations were confirmed.

    line 129
  19. Distinctive:

    Update inbox.md (action or archive pending items).

    line 130
  20. Distinctive:

    Never assume a metric. If the assistant does not know, it asks.

    line 153
  21. Distinctive:

    Never skip the check-in phase.

    line 154
  22. Distinctive:

    Keep each phase moving. If the user gives a one-word answer, accept it and move on.

    line 155
  23. Distinctive:

    If the user wants to skip a phase, let them.

    line 157
  24. Distinctive:

    All updates happen during the session, not after.

    line 158

“Common” is measured: kept by at least 6 in 10 of the 85 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Skill: Weekly Review

Trigger: "Let's do the weekly review", "weekly review", "weekly check-in"


What This Skill Does

A structured 30-60 minute session between the assistant and the user. The assistant asks, the user answers, the assistant updates. The goal: every context file, flag, and backlog stays current. No stale data. No assumptions. Any new session that reads snapshot.md after a weekly review has an accurate picture of where things stand.


Dependencies (Load Before Running)

Read these files before starting:

  1. personal/snapshot.md (current flags and last session)
  2. personal/day-ledger.md (this week's plan + debrief entries -- the primary data source for what happened)
  3. personal/patterns.md (behavioral observations with lifecycle states)
  4. .claude/rules/proactive-rules.md (guardrail tiers and proactive triggers)
  5. .claude/rules/evolution.md (self-check protocol)

Query these external sources during the review:

  • Task database: active tasks, completed this week, stale items (across all categories)
  • Calendar: past week events (what actually happened vs what was planned)

The 7-Phase Review

Run every phase in order. Each phase has specific questions the assistant asks the user. Do not skip phases. Do not rush. This is the one time per week everything gets audited.

Phase 1: Check-in (5 min)

Purpose: Get the human read on the week before looking at data.

The assistant asks:

  1. "How did this week feel? One word."
  2. "What was the win of the week?" (one thing that moved the needle)
  3. "What drained you the most?" (one thing that took more energy than it should have)

Listen. Do not optimize yet. This sets the emotional context for the rest of the review.

Phase 2: Metrics and Facts (10 min)

Purpose: Update things the assistant cannot verify on its own. Go through each one and ask directly.

The assistant asks about each metric it cannot track automatically:

  1. "Content/visibility: how many posts or outreach actions went out this week?" (update snapshot if needed)
  2. "Key conversations: any new ones since last review?" (update count)
  3. "Revenue: any money come in this week?" (update if changed)
  4. "Active leads: any movement?" (update flags)
  5. "Anything else happen this week I should know about?" (catch-all for things not mentioned in sessions)

Update snapshot.md flags with whatever the user reports. Remove resolved flags. Add new ones.

Phase 3: Backlog Audit (10 min)

Purpose: Query the task database and clean up what is stale.

Do this for each task category:

Primary Projects:

  • Query active tasks. Show the user the list.
  • Flag anything older than 14 days with no progress: "This has been sitting for [X] days. Keep, kill, or reprioritize?"
  • Flag any P1 tasks not in "Doing" status.
  • Check for duplicates.

Personal Life:

  • Query active tasks. Show the user the list.
  • Flag anything overdue or stale.
  • "Anything here that is done but not marked done?"

Work Tasks (if applicable):

  • Query active tasks. Show the user the list.
  • Flag overdue items.
  • "Any new items or issues this week?"

For each stale item, offer three options: (a) keep as-is, (b) reprioritize, (c) archive/close.

Phase 4: Calendar Review (5 min)

Purpose: Compare what was planned vs what actually happened.

  • Read this week's day-ledger.md entries first. If debriefs were committed, the plan-vs-reality data is already captured. Use it.
  • Pull last week's calendar events across all calendars to fill gaps (days without debriefs).
  • Estimate: how many hours went to each priority area?
  • Compare against the target time split (from goals or framework files).
  • If project hours are under target, name it: "You got [X] project hours this week. Target is [Y]. What got in the way?"
  • If work expanded beyond normal, name it: "Work took [X] hours this week. Is that temporary or a trend?"
  • Note which days had debriefs and which did not. If debrief coverage is under 50%, flag it: "Only [X] of 7 days had debriefs. The loop works best when it runs daily."

Phase 5: Patterns Check (5 min)

Purpose: Share behavioral observations and validate them with the user.

  • Review patterns.md. Are existing patterns still accurate?
  • Share any new observations from this week's sessions (things noted in snapshot.md flags as "watching").
  • If an observation has been confirmed twice, promote it to patterns.md.
  • If a pattern is no longer true, update or remove it.
  • Ask: "Anything you have noticed about yourself this week that I should know?"

Phase 6: Architecture Hygiene (5 min)

Purpose: Keep the assistant's files clean and current.

  • Check inbox.md: any pending items that should be actioned or archived?
  • Check snapshot.md: are all flags still relevant?
  • Check if any on-demand files were loaded in 5+ consecutive sessions (should they be promoted to always-loaded?)
  • Check if any always-loaded content was not used in recent sessions (should it move to on-demand?)
  • Flag any files that feel bloated or stale.
  • Ask: "Anything about how I work that felt off this week? Too noisy? Missing something?"

Phase 7: Next Week Setup (10 min)

Purpose: Set the priorities for the coming week.

The assistant asks:

  1. "What are the top 3 things that matter next week?" (the user decides, the assistant does not assume)
  2. "Any deadlines I should watch?" (add to flags)
  3. "Any people you need to follow up with?" (add to flags)
  4. "Anything you want me to proactively push you on?" (update proactive approach)

After the user answers:

  • Update snapshot.md with new flags and next-week priorities.
  • Update patterns.md if new observations were confirmed.
  • Update inbox.md (action or archive pending items).
  • Note the review date in snapshot.md: "Last review: [date]. Next: [date + 7 days]."

After the Review

  1. Summarize what was updated in one clean block:
    • Files updated: [list]
    • Flags added: [list]
    • Flags resolved: [list]
    • Tasks archived/closed: [count]
    • Patterns added/updated: [list]
  2. End with one observation. Something human. Not a task. A read on the week.
  3. Set the next review date. Suggest the same day/time next week.

Rules

  • The assistant asks. The user answers. The assistant updates. That is the flow.
  • Never assume a metric. If the assistant does not know, it asks.
  • Never skip the check-in phase. The human read matters.
  • Keep each phase moving. If the user gives a one-word answer, accept it and move on.
  • This is not a performance review. It is two people getting current. Warm, direct, efficient.
  • If the user wants to skip a phase, let them. Note what was skipped in the summary.
  • All updates happen during the session, not after. The user should see changes in real time.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.