Choose how each file is shown

Classify each file by its name and media type to choose how it is shown: as code, markdown, a diff, a picture, JSON or plain text, with the language, icon and label it is shown under

Give files ([{ path, mime_type? }]): each file's name or path and, when known, its media type, which wins over the extension. Answers classified, one per file in the order given: renderer (code, markdown, diff, image, json or raw), language, icon, label and the mime_type it was taken as. It reads only the names and types it is given: no file is opened, nothing is drawn, nothing is sent, and the signed-in classification owner the screen asks first is not called, so the answer is the file-content renderer's deterministic fallback. Refuses an empty list or a file without a path. The screen (show/FileContentRenderer.tsx and the media player) also draws the file, plays sound and video, pages PDFs and edits the facts kept about a file; this run does none of that.

What it asks for

Nothing: your AI starts it with your words.

Use it from your AI

Connect the AI you already use to Trillion once. Then ask in your own words; your AI finds this tool by what it does and runs it.

Say to your AI: “Use Trillion to classify each file by its name and media type to choose how it is shown: as code, markdown, a diff, a picture, JSON or plain text, with the language, icon and label it is shown under”