technique reference · by file type
Hidden text in Excel spreadsheets
A spreadsheet can hide text in cells formatted to look empty, in white font on a white fill, in hidden rows, columns and sheets — including sheets Excel's own menus cannot unhide — and in comments and print headers.
Anything that reads the workbook returns that text in full, so an AI model analysing the file reads cells nobody looking at the grid could see. Below is every way we detect hidden text in Excel files.
xls.styles-not-readA scope disclosure, not a concealment technique: it says out loud that a legacy binary .xls has its text read and its styling and sheet visibility not read.xlsx.blank-number-formatA custom number format made only of empty sections renders the cell visually empty while its value survives in the file in full.xlsx.clipped-textA wrapped cell holding more lines than its row can show draws the first line and nothing after it; a parser reads every line.xlsx.commentComment text lives in its own part of the workbook, shows the grid only a small marker, and is read in full by anything walking the package.xlsx.conditional-blank-formatThe same blank-rendering number format, delivered by a conditional-formatting rule instead of the cell's own style, so the cell's static formatting looks entirely ordinary.xlsx.hidden-row-colA row or column that is hidden — by the flag, or by being collapsed to nothing — shows a reader an unbroken grid while its cells read out normally.xlsx.hidden-sheetA worksheet marked hidden keeps every cell it holds, and the text is read out of the package as if the sheet were open.xlsx.number-format-substitutionA number format whose first section is a quoted literal paints those words over every value, so the grid says one thing and the stored cell says another.xlsx.offrange-contentText placed outside the range a sheet declares as used — by one column or by four thousand rows, the engine does not distinguish — is somewhere nobody scrolls, and is read in full by anything that walks the sheet.xlsx.print-header-footerText in a sheet's print header or footer appears on paper and in Page Layout view, and never in the Normal view a reviewer scrolls.xlsx.tiny-fontA cell whose font is set to a sub-readable point size occupies the grid but shows a reader nothing, while its text extracts at full length.xlsx.very-hidden-sheetA worksheet in the `veryHidden` state cannot be reached through Excel's interface at all, and its cells are still read in full.xlsx.white-fontA cell whose font colour matches the fill behind it reads as an empty cell on screen and as its full contents to a parser.
shared by every Office file
ooxml.embedded-font-remapAn embedded font maps a digit to a glyph that cannot be that digit, so a reader sees different numbers from the ones the document stores.ooxml.raster-instructionAn embedded image was read and the text inside it is shaped like an instruction aimed at a model.ooxml.uninspected-embedded-objectA document embedded inside this one could not be opened, so whatever it contains was not analysed.ooxml.uninspected-fontThe document embeds a font whose glyph mapping this engine could not read, so the extracted text may differ from what a reader sees.ooxml.uninspected-partA part inside the OOXML package carries text that the format's parser never opened, so its contents were not analysed.ooxml.uninspected-rasterThe document carries an embedded image large enough to hold readable text that this engine did not read.