agentsclimarketplace

Spreadsheet basics

Skill protosphinx/sphinxstack/skills/spreadsheet-basics

A stack of skills for your AI agent, made for people starting from zero. 103 skills and 149 project briefs — load one into Codex, Claude Code, Copilot, Gemini, or Cursor and do the thing.

Install
npx -y skills add protosphinx/sphinxstack --skill spreadsheet-basics

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 18 days oldThe repository was created 18 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Build spreadsheet fluency on data they already track. Structure, formulas, lookups, and summaries on their own sheet. Use when they say "teach me spreadsheets", "help me with Sheets/Excel", or bring data living in their notes app.

SKILL.md

3.0 KB, 642 tokens by cl100k_base, as published. Nobody here has run it

spreadsheet-basics

Teach spreadsheets to someone who has probably only seen them as grids other people fill in. The way in is their own data: something they already track in a notes app, a group chat, or their head. Use Google Sheets or LibreOffice Calc, whichever they have. Every formula you introduce answers a question they actually asked.

Find the data first

Ask what they keep track of, even loosely: grades and assignments, match results for a game they play, money in and out, workouts, a collection, hours at a job. Pick the one with the most rows in it. Ten real rows beat a hundred invented ones — never make up demo data when their own exists. If the data is on paper or in messages, entering it together is the first lesson: it forces the structure conversation.

Structure before formulas

Build the sheet with them, talking through each choice:

  • One row per thing that happened, one column per fact about it. Header row, frozen. No merged cells, no color-as-data.
  • Each column holds one type: dates formatted as dates, numbers as numbers. Show what breaks when "3 wins" sits in a number column.
  • If two distinct entities genuinely exist (games and players, purchases and products), give each its own tab. Otherwise keep one data table and avoid structure that the real data does not need.

Formulas that answer their questions

Introduce each function the moment their question needs it, never as a list to memorize:

  1. Arithmetic and cell references — a total or difference column. Fill down; show that the reference moves.
  2. SUM, AVERAGE, COUNT, MIN, MAX over a whole column.
  3. IF for a judgment column: pass/fail, over/under budget, win/loss.
  4. COUNTIF and SUMIF — "how many losses?", "how much on food?"
  5. If the data has a real second entity, use a lookup to pull one of its facts. Otherwise use another formula the person's question actually requires.
  6. A summary table: one block, built with UNIQUE + SUMIF or a pivot table if the tool has one — totals per category, per month, per opponent. This is the payoff: the sheet now says something the raw rows did not.

At each step, have them type the formula. Predict the result out loud before pressing enter; wrong predictions are the best teaching moments.

Done

  • A sheet of their real data with a clean header, typed columns, and only the tabs its structure requires
  • Working formulas they wrote themselves, including COUNTIF or SUMIF and a lookup when the data genuinely requires one
  • A summary block that answers a question they cared about, with the answer stated in a sentence at the top of the sheet
  • They can add a new row and watch every summary update

Then: clean-a-dataset when they meet messier data, or chart-the-truth to turn the summary block into a chart.

What ships with it: 1 file

217 B alongside SKILL.md

agents/

Gives 0 of the 12 instructions most pdf office docs skills give in 642 tokens

Counted across 636 of the 690 authors here whose files we hold, read 2026-08-07

  • Extract text or tables using pdfplumber or pdftotextin 89 of 636, across 23 files
  • Create new PDFs using reportlabin 83 of 636, across 16 files
  • Read forms.md before filling out PDF formsin 80 of 636, across 13 files
  • OCR scanned PDFs using pytesseract and pdf2imagein 77 of 636, across 10 files
  • Use qpdf to merge or split PDFs or large filesin 70 of 636, across 3 files
  • Use Excel formulas instead of hardcoded calculated values or Python calculationsin 68 of 636, across 13 files
  • Unpack, edit, and repack XML for existing documents or presentationsin 63 of 636, across 8 files
  • Document sources for all hardcoded valuesin 61 of 636, across 9 files
  • Write minimal, concise Python code without unnecessary commentsin 59 of 636, across 7 files
  • Run the recalculation script (recalc.py) after adding or modifying formulasin 59 of 636, across 7 files
  • Fix all identified formula errors and recalculate before finishingin 58 of 636, across 6 files
  • Format years as text stringsin 57 of 636, across 5 files

Said here and by no other author read

  • use their existing personal data
  • never invent demo data
  • pick the dataset with the most rows
  • enter messy data together first
  • use one row per event and one column per fact
  • avoid merged cells and color-as-data

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.