Contenu du pack

Un fichier SKILL.md (le guide pratique) + un modèle de départ. Aperçu :

Activity Calendar Chart

When to use

• You want to see a year of habits, commits, workouts, or sales as a single picture instead of a spreadsheet
• You need a "year in review" visual that shows streaks, gaps, and seasonal patterns at a glance
• Your team wants a shared view of activity — support tickets, deploys, posts shipped — across weeks and months
• You are reporting progress to a client or manager and a heatmap says more than a table of daily counts
• Your raw data is a list of dated events and you need it turned into something a reader can understand in five seconds

Workflow

1. Define the unit (input: raw dated events → output: counting rule)
   Decide exactly what one cell means before charting anything: one day = one cell, and the value is a count (workouts done, tickets closed) or a yes/no (published or not). Write the rule down: "one cell = one day, shade = number of focused work sessions that day." Ambiguous counting produces a chart that looks precise and means nothing.
2. Clean the dates (input: event list → output: tidy daily table)
   Normalize everything to a single daily table: one row per day in the range, zero-filled for days with no activity. Check for the classic traps — timezone shifts, duplicate entries, events dated in the future, and ranges that start mid-week. A heatmap with missing zeros lies: an empty day and an unrecorded day must be distinguishable or the gap pattern is meaningless.
3. Choose the scale (input: daily table → output: color scale + legend)
   Pick a scale that matches the data: a sequential scale (light to dark) for counts, a binary scale for yes/no. Set the breakpoints from the actual distribution — if most days are 0–2 and a few are 20, a linear scale will wash everything out; use thresholds that separate the common range. The legend must state the unit: "darker = more sessions, max 8/day."
4. Annotate the story (input: scaled chart → output: annotated chart)
   Add the annotations that turn a grid into a story: label the streaks, mark the gaps with a reason you actually know ("travel", "holiday"), and call out the peaks. Keep annotations short and place them outside the grid where they cannot obscure cells. Never annotate a gap with a reason you are guessing — an unexplained gap is honest; an invented explanation is fabrication.
5. Check readability (input: annotated chart → output: publish-ready chart)
   Verify the chart at the size it will be read: month labels aligned, weekday labels legible, legend visible, color differences distinguishable without the legend. Check it in grayscale — if the scale collapses when color is removed, adjust the lightness steps. Confirm the total shown matches the source data (sum of cells = total events) before publishing.

Aperçu du modèle

Activity Calendar Chart — Data & Chart Template

One dataset = one chart. Define the counting rule and clean the dates before shading a single cell.

Dataset: {what is being tracked}

Counting rule (one cell = one day, value = ?): {e.g. "number of focused work sessions that day"} Date range: {start} → {end} Timezone: {timezone used for all dates} Source of events: {where the event list came from}

Data cleaning checklist

[ ] One row per day in range, zero-filled for days with no activity [ ] Timezone shifts checked (no events landing on the wrong day) [ ] Duplicates removed [ ] Future-dated events removed or flagged

Le pack de compétence lui-même (SKILL.md + modèle) est rédigé en anglais.

Acheter le pack complet — $4.99 Pro — les 511 compétences