xlsx

name

xlsx

description

Generate or edit real Excel (.xlsx) spreadsheets with formatting — colored cells, fills, bold headers, borders, number formats, column widths, multiple sheets, simple formulas. Runs `openpyxl` via `uv`, so nothing is installed globally and it always works even after a cache clean. Saves into the current topic folder and sends via the bot MCP `send_document` tool. Also applies when the user sends an existing .xlsx and asks to change or restyle it. Use when: "сделай эксель", "сделай таблицу в экселе", "эксель с цветами", "excel с раскраской", "сделай xlsx", "покрась ячейки", "сделай .xlsx", "таблица с форматированием", "поправь эксель", "make an excel", "make an xlsx", "excel with colors", "colored cells", "formatted spreadsheet", "edit this xlsx". Prefer this over CSV whenever the user wants colors, formatting, bold headers, multiple sheets, or anything CSV can't carry. CSV is only the fallback when the user explicitly wants a plain, library-free table.

xlsx

Generate .xlsx files with openpyxl, run via uv so nothing is installed globally. Save into the current working folder, then send the file back through the bot MCP.

Do not fall back to CSV just because "the openpyxl env is gone." It is never gone. uv run --with openpyxl re-materializes openpyxl from PyPI on demand — even right after uv cache prune, even if no python env has it installed. If you ever hit ModuleNotFoundError: openpyxl, you invoked the wrong python — use the uv run --with openpyxl command below, never bare python3/python or the repo's .venv. CSV is only correct when the user explicitly asks for a plain library-free table, not as a workaround.

Workflow (create)

1. Clarify briefly only if the request is thin: what columns, is there sample data, does it need colors/conditional formatting, one sheet or several. One short message, max 2–3 questions. Don't interrogate.
2. Generate the script as make_xlsx.py in the current folder, from the template below. One script per workbook — easier to re-run after tweaks.
3. Run with uv, no venv needed:

   uv run --with openpyxl python make_xlsx.py
   

4. Send the file via bot MCP send_document with the resulting .xlsx. Short caption in Russian saying what's inside and offering edits.
5. Keep the script and xlsx in the topic folder — user may ask to tweak.

Read-edit-verify loop (edit)

If the user sent a .xlsx or referenced one in the folder:

1. Read — wb = openpyxl.load_workbook(path), iterate ws.iter_rows(values_only=True), dump a flat plain-text preview of each sheet. Show it before editing so the user confirms the target.
2. Edit — mutate cells (ws.cell(row, col).value = ...), save as edited.xlsx. To keep existing formatting, load with the default (not read_only) and don't rebuild sheets you aren't changing.
3. Verify — reopen edited.xlsx, confirm it opens without exception and the expected sheets/row counts survived. If verify fails, say so — don't send a broken file.

Colored-cells template (make_xlsx.py)

from openpyxl import Workbook
from openpyxl.styles import Font, PatternFill, Alignment, Border, Side
from openpyxl.utils import get_column_letter

wb = Workbook()
ws = wb.active
ws.title = "Данные"

headers = ["Задача", "Статус", "Приоритет"]
rows = [
    ["Собрать отчёт", "Готово",   "Высокий"],
    ["Проверить данные", "В работе", "Средний"],
    ["Отправить клиенту", "Ждёт",    "Низкий"],
]

# Header style: bold white on dark fill, centered, with border
header_fill = PatternFill("solid", fgColor="1F4E78")
header_font = Font(bold=True, color="FFFFFF")
center = Alignment(horizontal="center", vertical="center")
thin = Side(style="thin", color="BFBFBF")
border = Border(left=thin, right=thin, top=thin, bottom=thin)

for c, title in enumerate(headers, start=1):
    cell = ws.cell(row=1, column=c, value=title)
    cell.fill, cell.font, cell.alignment, cell.border = header_fill, header_font, center, border

# Status -> color mapping for conditional-style fills
status_fill = {
    "Готово":  PatternFill("solid", fgColor="C6EFCE"),  # green
    "В работе": PatternFill("solid", fgColor="FFEB9C"),  # yellow
    "Ждёт":    PatternFill("solid", fgColor="FFC7CE"),  # red
}

for r, row in enumerate(rows, start=2):
    for c, val in enumerate(row, start=1):
        cell = ws.cell(row=r, column=c, value=val)
        cell.border = border
        if c == 2:  # Статус column
            cell.fill = status_fill.get(val, PatternFill())
            cell.alignment = center

# Auto-ish column widths from content length
for c, title in enumerate(headers, start=1):
    width = max(len(str(title)), *(len(str(row[c - 1])) for row in rows)) + 4
    ws.column_dimensions[get_column_letter(c)].width = width

ws.freeze_panes = "A2"  # keep header visible when scrolling
wb.save("output.xlsx")
print("saved output.xlsx")

Extend as needed: wb.create_sheet("Sheet2") for multiple tabs; cell.number_format = "#,##0.00 ₽" for money; cell.value = "=SUM(B2:B4)" for formulas; openpyxl.formatting.rule for real conditional formatting.

What not to do

• Don't use bare python3 / python or the repo .venv — always uv run --with openpyxl python make_xlsx.py.
• Don't fall back to CSV to dodge a ModuleNotFoundError — fix the command instead.
• Don't invent data to fill cells; ask for the source if you don't have it.
• Don't silently send the file if the verify step failed.