quarterly-recap
Skill2 starsChanged 48 days agoMITView on GitHub
Write the end-of-quarter review from that quarter's three monthly recaps, then check config/season.md against what the quarter actually showed and propose an amendment or a rewrite where it no longer fits. Use when the user asks for a quarterly recap, a quarter in review, how the quarter went, or to close out Q1, Q2, Q3, or Q4. Writes reviews/YYYY-QN-quarterly.md and publishes it to the Recaps artifact page.
In this repository7 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- annual-reviewSkill
- captureSkill
- daily-briefingSkill
- daily-wrapSkill
- monthly-recapSkill
- weekly-recapSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
25 rules. About 0 are common, 16 are distinctive, and 9 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈16
Distinctive · estimate
This author's own choices. Worth a look.
9
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Distinctive: line 26
Never read a weekday off an ISO string.
- Very specific: line 33
If config/profile.md carries no timezone, stop and ask the user for theirs rather than guessing or reading the machine zone.
- Distinctive: line 39
If the quarter has no monthly recaps and no weeklies either, say so and stop.
- Very specific: line 42
If reviews/YYYY-QN-quarterly.md already exists for the quarter you are about to write, stop and ask before replacing it.
- Distinctive: line 127
Do not re-run any sweep. Never invent content for a missing source.
- Very specific: line 135
Read config/season.md for the lens, and take the life threads from config/profile.md.
- Distinctive: line 140
If the evidence says the season named there no longer matches how the quarter was actually lived, that is the central finding of the …
- Distinctive: line 158
Every slot still unfilled, nothing written anywhere: the user has not been set up.
- Distinctive: line 170
Write the sections in this order.
- Distinctive: line 175
Always first, and never omitted, because everything after it depends on how much of the quarter was actually readable.
- Distinctive: line 265
If the main thing barely moved in three months, say that plainly.
- Very specific: line 274
If config/season.md question 4 names non-negotiables, read the floor against them explicitly and say whether it held across ninety days, …
- Very specific: line 283
Only when
config/season.mdquestion 5 names one: If that answer is blank or marked (skipped), omit this section and never mention it. - Distinctive: line 302
If the patterns did not move this quarter, say so in one line.
- Distinctive: line 324
Start with the read: what the season file says this stretch is for, versus what the three months actually show.
- Distinctive: line 343
Write any draft in the user's own language wherever the sources give it to you.
- Distinctive: line 344
Keep every proposed answer short enough to be answerable, and hold the same punctuation rules as the rest of the review.
- Distinctive: line 357
Keep the current season, in which case say what evidence they are choosing against, once, and drop it.
- Very specific: line 360
Write
config/season.mdonly after the user says yes: Not when the case looks obvious, not when they say the draft reads well, not when … - Very specific: line 386
Write to reviews/YYYY-QN-quarterly.md, for example reviews/2027-Q3-quarterly.md.
- Very specific: line 397
Add the user's own tags from config/profile.md alongside #meta where they fit the quarter.
- Distinctive: line 423
Commit, staging the file by its own path:
- Distinctive: line 430
Never git add -A. Any season rewrite the user approved is a separate commit, as above, and unrelated work sitting in the tree is not yours …
- Very specific: line 457
No em dashes and no double hyphens in the prose of the review you generate, and none in a proposed season file either.
- Distinctive: line 462
Write numbers as digits. This is a written review, and "sixty-two" reads stilted on the page where "62" does not.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Quarterly Recap
You are writing an end-of-quarter review built on the quarter's three monthly recaps. The job is to find the arc no single month could see: where the quarter opened versus where it closed, which decisions held and which got walked back, which patterns hardened into something real. Then you propose what the next season should be.
When this runs
At the end of a calendar quarter, or in the first days of the next one. The user can also ask for a past quarter at any time.
Confirm which quarter before reading anything. Q1 runs January first through March
thirty-first, Q2 April first through June thirtieth, Q3 July first through
September thirtieth, Q4 October first through December thirty-first. Compute the
boundaries and the run weekday through Bash, with the timezone from
config/profile.md set on the command itself: TZ=Europe/Lisbon date +%Y-%m-%d
and TZ=Europe/Lisbon date +%A, substituting the user's actual zone. A bare date
reads the machine clock, which is not necessarily the clock the user lives on.
Never read a weekday off an ISO string.
Backing up from today is not portable between systems, so use the form that
matches the machine: TZ=Europe/Lisbon date -v-3m +%Y-%m on macOS,
TZ=Europe/Lisbon date -d '3 months ago' +%Y-%m on GNU. Try one and fall back to
the other when it errors, and keep TZ= on whichever you end up using.
If config/profile.md carries no timezone, stop and ask the user for theirs
rather than guessing or reading the machine zone. Setup collects it, so this
should not happen, and if it has then something is wrong with the file.
Title the review by quarter and year, for example "Q3 2027".
If the quarter has no monthly recaps and no weeklies either, say so and stop. A quarterly built on nothing is fiction.
If reviews/YYYY-QN-quarterly.md already exists for the quarter you are about to
write, stop and ask before replacing it. Say when it was written and what it was
built from, which monthlies or weeklies backed it, how many of those carried
coverage: thin, and whether it is marked thin itself. Then let the user choose:
replace it, write this one under a different name, or cancel. A quarterly
assembled from three full monthlies should not be silently overwritten by a run
working from two weeks of weeklies.
Sources marked thin
A weekly carries coverage: thin when it was built from fewer than three daily
notes, and a monthly carries it when half its weeks or more came in thin. Check
that field on every recap you read, before you count anything.
Weight a thin source as partial. It does not carry a trend on its own, and it is never the sole support for a win, a gap, a decision, or a movement read across ninety days. Where a thin month is all you have for a stretch, say what the stretch rests on rather than writing around it.
Thin months also change what the quarter can claim. Three months where two are
thin is not ninety days of evidence, so say that near the top of the review and
keep The Quarter in the Rearview and How the Quarter Moved proportionate to it.
Carry coverage: thin into this review's own frontmatter when half the quarter
or more came from thin sources, so the annual sees it. It also weakens the case
for a Season Rewrite, so hold that section to what the sources actually support
and say plainly when the evidence is too thin to propose one.
Sources (read in this order)
reviews/YYYY-MM-monthly.mdfor each of the quarter's three months. These are the spine of the whole review.For any month with no monthly recap, fall back one rung to that month's
reviews/YYYY-MM-DD-weekly.mdfiles, using the monthly skill's straddle rule to decide which weeks count as that month's. Record it in Source Coverage and Gaps: which month lacked a monthly, which weeklies you read instead, and what dates those weeklies actually cover. Do not paper over it, and do not drop to daily notes to fill it.A month with neither a monthly nor any weeklies is an unrecoverable gap. Say so directly, name the dates with no coverage under them, and stop there. Do not reconstruct it from daily notes, and do not let it be quietly absorbed into the quarter's story as though it had been read. A quarter missing a month is a two-month review of a three-month period, and it has to read that way.
tracker.mdfor what stalled across ninety days. An item that survived two or three separate monthly stale lists is stronger signal than one that went stale once. The Completed archive shows what actually closed.patterns/for which patterns were promoted, strengthened, or appended across the quarter, and which watched items resolved or fizzled.projects/decision records, and only those, for the Decisions of the Quarter section. This is the one read that reaches below your rung. See the note below for why it is allowed and how narrow it is.config/season.mdandconfig/profile.mdfor the lens and the thread list.
Never re-crawl daily/. Three rungs of reduction sit below you already, the
evening wrap, the weekly, and the monthly, and re-reading raw days throws that
work away.
Two of the reads above are lateral, not downward. tracker.md and patterns/ are
running records that every rung reads directly, because neither is a summary of
the rung below it. Reading them here is normal and is not an exception to
anything.
Running records show today, not quarter end. Pattern files keep getting appended after the quarter closes, the tracker keeps changing, and a review run three weeks into the next quarter reads all of that as if it were the state on the last day. Nothing in the files warns you, and it gets worse the later the review runs. So:
- Filter every pattern entry to the quarter's dates and ignore anything later. Otherwise a pattern looks like it hardened inside the quarter when it did not.
- Where the recaps and a running record disagree about the state at quarter end, the recaps win. Mark anything dated after the quarter as later movement rather than folding it into the arc.
- The tracker keeps no history of how long an item sat before it closed, so do not claim a duration it cannot support.
Pattern files quote daily notes directly, so daily-note content can reach this
review second-hand. Using it that way is fine; label it as indirect when you do.
Opening daily/ yourself is still off limits.
projects/ is the real exception, so hold it to its scope: decision records
only, for Decisions of the Quarter. A decision written straight into a project
file can go three months without ever reaching a weekly or a monthly, and that is
the specific hole this read closes. Do not read project status, history, or notes
while you are in there.
Do not re-run any sweep. Never invent content for a missing source. If a month was thin, say it was thin.
Connectors are not part of this pass. Everything a calendar or an inbox had to say reached the daily and weekly rungs months ago.
Current season
Read config/season.md for the lens, and take the life threads from
config/profile.md. Use the threads the user named there, in their order. Never
assume a thread they did not write down.
Ninety days is long enough that the season may have shifted underneath the file. If the evidence says the season named there no longer matches how the quarter was actually lived, that is the central finding of the review, and Season Rewrite is where it goes.
When the config is not fully filled
Judge the answer slots, not the file text. Both config files carry instruction comments that describe the placeholder markers in prose, and those comments stay in place on a fully configured vault, so searching for that shape matches every time and tells you nothing. Look at what sits under each numbered heading.
A slot counts as filled when it holds anything the user put there, and
(skipped) is something the user put there. It is a deliberate pass on an
optional question, so read it as answered and move on. A skipped season question
never blocks this review.
Then degrade rather than stop:
- Every slot still unfilled, nothing written anywhere: the user has not been set up. Say so in one line, offer the setup interview in CLAUDE.md, and stop.
- Some filled, some not: run the review on what is there. No season means no season lens, so drop Season Lens, report The Honest Score against the profile's life threads instead, and say nothing about the missing file. Season Rewrite becomes an offer to write a season for the first time rather than a proposed replacement, and it still needs an explicit yes before anything is written.
A partly filled config is a normal state, not an error.
What to include
Write the sections in this order. If a section has nothing real to say, leave it out rather than padding it.
Source Coverage and Gaps
Always first, and never omitted, because everything after it depends on how much of the quarter was actually readable. When coverage is complete, it is one line saying so.
Otherwise name, month by month, which had a monthly, which fell back to weeklies (and which ones), which came in thin, and which are unrecoverable. Then state the dates genuinely covered and the dates that are not. Close with a plain read on how much of the ninety days has a recap under it and how much weight the rest of the review can bear.
Season Lens
One or two bullets naming the lens this review is read through, from
config/season.md. This sets what counts as meaningful across three months.
The Quarter in the Rearview
Five to seven bullets on the arc and texture of the quarter across every thread. Where it opened, where it closed, what it actually felt like to live through. Not a chronological replay of three months. The version a friend gives after reading everything.
How the Quarter Moved
One bullet per thread in config/profile.md. Start-state versus end-state,
direction, and distance. This is the section that earns the quarterly its place.
Name the movement a single month cannot see, including the slow drifts that only
register at ninety days. A thread that went untouched for three months gets a
bullet saying exactly that.
Wins of the Quarter
Stack-ranked, most impactful first. Elevated and de-duplicated from the monthly wins, keeping only what still matters at quarter altitude. A win that felt large in the first month and left no trace by the third does not belong here. Write each as a natural phrase, not a trophy label. Skip the section rather than manufacture a weak win.
Decisions of the Quarter
Harvested from the monthlies' Decisions of the Month sections, from decision
records in projects/, and from the tracker's Completed archive, where an item
closed with a stated reason is a decision record and often the only one. Treat
the three as sources of equal standing. Four to ten items total, grouped four
ways. Drop any group that is empty.
- Held, meaning decided and still standing at quarter end. One line each on what it bought or what it cost.
- Walked back, meaning decided and then reversed or quietly abandoned. Say what changed where the sources support it. Flag reversals plainly rather than softening them.
- Still open, meaning named as a decision to make and still unmade. Note how long it has been sitting, in months if that is the honest unit.
- Decision debt, meaning choices being made by default because nobody made them on purpose. These are the expensive ones, and also the easiest place in this review to invent something. Every decision debt item names its evidence trail: the specific recap lines or stalled tracker items that show the default being taken. An item you cannot attach evidence to gets dropped, not softened into a hedge.
A decision that appeared as held in month one and walked back in month three belongs in Walked back, with the arc stated in the line.
Gaps
The matched counterpart to Wins. Honest self-assessment across the quarter: judgment calls, dropped balls, avoidance, missed follow-ups, habits worth correcting. Use stalled-thread evidence as support, especially tracker items that went stale in two or three separate months, but frame it as performance rather than circumstance. Honest, not harsh. Skip it if there is genuinely nothing real.
<!-- Note for the composer: Gaps is a this-quarter self-assessment. It is
point-in-time and it can be true the first time it happens. Patterns, further
down, is cross-time recurrence, threshold-gated and neutral. A gap that recurred
all quarter may also appear as a hardened pattern below. Naming it in both, from
each section's angle, is correct. -->The Main Thing
config/season.md names what this season is primarily about. At quarter altitude
this section tells the full arc of that one thing, with named specifics
throughout.
Cover what entered over the three months and where it came from, what advanced and how far it got, what died and at what stage, what went quiet without ever formally ending, and what is live at quarter end with its next step. Then say what shape the quarter had. Ninety days is long enough to show where the trouble sits: at the start of things, in the middle, or at the finish. If the same failure point repeats across several attempts, name it.
If the main thing barely moved in three months, say that plainly. It is the most useful sentence in the section.
Trend Lines
Energy and mood across the quarter, plus any rhythm the notes actually track. Describe the shape across three months: steady, climbing, sawtooth, a recovery that held or did not. Tie shape to cause only where the sources support the link.
If config/season.md question 4 names non-negotiables, read the floor against them
explicitly and say whether it held across ninety days, using the monthly floor
reads rather than recounting days yourself. If the season was rewritten mid-quarter
and the non-negotiables changed with it, read each one against the stretch it
actually covered. If that answer is blank or marked (skipped), skip the floor read
entirely and do not mention it.
The Evening Question
Only when config/season.md question 5 names one. If that answer is blank or
marked (skipped), omit this section and never mention it.
A quarter holds roughly ninety answers to the same nightly question, which is the longest run of one signal anywhere in this folder. Say what the answers show over that span: what held, what shifted and roughly when, and whether the question is still pulling anything useful. Two to four bullets. If the answers have gone thin or repetitive, say so plainly and note that Season Rewrite is where the question gets changed or dropped.
Patterns: What Hardened, What Faded
Consolidation over patterns/. Review only, and do not run the pattern sweep.
Cover which patterns were promoted or strengthened this quarter with links to the
files, which watched items became real patterns and which fizzled, and retirement
candidates. A pattern with no confirming signal in a full quarter has probably
stopped being true, so name it and recommend retiring it. Recommend only. Do not
edit or delete a pattern file.
If the patterns did not move this quarter, say so in one line.
People and Relationships
Who mattered across the quarter, who got invested in, who went quiet, who deserves
a reconnect or a thank-you. Quarter altitude, so only the names with weight. One
line each on why they surfaced and what the nudge is. Three months of silence from
someone who was central in the first month is worth naming. Use [[Name]] links.
The Honest Score
A candid quarter-level read against what the season calls for. Not a grade, not a pep talk. What got leaned into, what got let slide, whether ninety days moved things toward what was actually wanted. If the quarter was rough, say so plainly and do not soften it into a lesson.
Season Rewrite
This is the headline ritual of the quarterly. A quarter is the right interval to
ask whether config/season.md still describes real life, and the answer comes
from evidence rather than from how the user feels on the day they run this.
Start with the read: what the season file says this stretch is for, versus what the three months actually show. Attention allocation is the sharpest test. If the file says a third of the user's energy goes to one thread and the quarter shows it went untouched for six weeks, that gap is the finding.
The answer is rarely all or nothing, so land on one of three:
- The season still fits. Say so in a line and skip the draft.
- One or two answers are wrong. This is the common case. A quarter can be
lived exactly as the season describes while the attention split or a
non-negotiable turns out to be set at the wrong level. Quote the specific
answers in
config/season.mdthat no longer match, draft replacements for those answers only, and leave the rest of the file alone. - The season no longer describes the quarter at all. Draft a full replacement
config/season.mdinside the review, following the five questions in the current file: what the season is about, where attention should go, what gets tracked without being asked, the non-negotiables, and the optional evening-wrap question.
Write any draft in the user's own language wherever the sources give it to you. Keep every proposed answer short enough to be answerable, and hold the same punctuation rules as the rest of the review.
The Evening Question section above is the evidence for the fifth answer. Keep the question if ninety nights of answers are still pulling something, swap it if the answers went flat, and leave it blank if it never earned its place. Do not propose a new nightly question with no read behind it.
Then walk the user through whatever you proposed, one choice at a time:
- Accept it as written.
- Edit it, in which case ask which answers they want to change and work through only those.
- Keep the current season, in which case say what evidence they are choosing against, once, and drop it.
Write config/season.md only after the user says yes. Not when the case looks
obvious, not when they say the draft reads well, not when they go quiet. An
explicit yes, then write the file and commit it on its own, separately from the
review:
git add config/season.md
git commit -m "season: updated after YYYY-QN"Carry Into Next Quarter
Three to five items, maximum. What genuinely deserves early attention: time sensitive, deferred too long, or unusually high impact given the season. Not a to-do dump. If something has been carried forward from two prior monthlies without moving, either say why it is still here or recommend dropping it.
One Thing to Sit With
A single closing line. A quote from the user's own words, a pattern that deserves more reflection than a bullet holds, or an observation worth carrying into the next ninety days. Something a friend who was paying real attention would notice, not a generic affirmation.
Output
- Write to
reviews/YYYY-QN-quarterly.md, for examplereviews/2027-Q3-quarterly.md.
---
type: review
date: YYYY-MM-DD # last day of the reviewed quarter
tags: ["#meta"]
---Add the user's own tags from config/profile.md alongside #meta where they fit
the quarter. Add coverage: thin when half the quarter or more came from thin
sources, using that exact lowercase value, since the annual looks for that string.
- Title the body
# Quarterly Recap: <Quarter> <Year>. - End with a sources footer, in the same shape the weekly and monthly use, wikilinking the monthly recaps you read plus the substitute weeklies for any month that had no monthly, and marking any of them that came in thin:
---
**Sources:** [[2027-07-monthly]], [[2027-08-monthly]], [[2027-09-06-weekly]],
[[2027-09-13-weekly]], tracker.md- Publish it to the Recaps page, following "
recaps/<doc id>" inartifacts/README.md, with doc idquarterly-YYYY-QN. Readrecaps_urlfromconfig/artifacts.mdfirst:pages: off, skip this step; blank in an interactive session, run "Setup" in that README for the Recaps page; blank and unattended, skip it and sayArtifact: skipped (not set up yet); no Artifact tools in the session, skip it silently. Build the doc from the file you just wrote, attach atracker_lineonly where you know the exact tracker line an item came from, write it totmp/recap.json, thenArtifactDatasetintorecaps. The page shows the Season Rewrite section as plain text; the decision aboutconfig/season.mdstill happens here in the conversation. A failure never blocks the commit: noteArtifact: failed (<reason>)and carry on. - Commit, staging the file by its own path:
git add reviews/YYYY-QN-quarterly.md
git commit -m "quarterly-recap: YYYY-QN"Never git add -A. Any season rewrite the user approved is a separate commit, as
above, and unrelated work sitting in the tree is not yours to commit.
Then check for a remote. Push only to a remote the user owns. If there is no remote, or the only one belongs to somebody else, stop after the commit and say nothing about it.
- Confirm with a short summary and the output path, and state where the season question landed: still fits, accepted, edited, or kept. Do not read the review back at the user.
Tone and format rules
Honest, warm, specific. Someone who read three months of notes and is giving back the real version. Not a performance review, not a productivity report, not a coach with a framework.
Length lands around eight to ten minutes of reading, a notch above the monthly because a quarter earns more depth. The extra length belongs in Decisions, The Main Thing, and Season Rewrite rather than spread evenly. Longer is not better.
Bullets are the default for every section. The only prose section is the closing line, though the Season Rewrite draft follows the shape of the season file itself. Even the longer sections are bullets, written as short sentences with natural connective phrasing so they read like a person talking. Keep any unavoidable paragraph to three or four sentences. Use whitespace between sections.
No em dashes and no double hyphens in the prose of the review you generate, and
none in a proposed season file either. Use commas, parentheses, or a new
sentence. It is a rule about prose, so markdown --- rules and the frontmatter
delimiters are fine.
Write numbers as digits. This is a written review, and "sixty-two" reads stilted on the page where "62" does not. Dates and weekdays keep their normal form.
Avoid:
- Filler and flattery. No "what a quarter", no "you crushed it", no "you should be proud".
- Filler openers like "here is your quarterly summary". Start on substance.
- Using the user's name more than once.
- Summarizing the three months in order. This is a synthesis across the monthlies, not a replay of them.
- Sections that exist to fill space.
- Turning The Honest Score into encouragement.
- Manufacturing a win, gap, decision, pattern, or insight the source recaps do not clearly support.
- Writing to
config/season.mdwithout an explicit yes.
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
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.

