# Setup

ClearRead is designed for a current Omarchy Quattro installation. A standard
installation already contains its runtime pieces: the Omarchy capture helper,
Hyprland tools, `grim`, `tesseract`, English Tesseract data, `wl-copy`,
`wl-paste`, Python, and `setpriv`.

## Check readiness

From the installed plugin directory:

```bash
./bin/clearread doctor
```

The command prints one JSON object. `ready: true` means at least one reading
action is available. Inspect `capabilities` for the active-window, region,
clipboard, and explicit-copy paths; one unavailable command does not hide the
rest of the reader.

The doctor does not capture a window, open the region picker, read the
clipboard, or run Tesseract recognition.

## OCR languages

English is the default:

```bash
tesseract --list-langs
```

Before launching the Omarchy shell, set one or more lowercase Tesseract
language codes made only of ASCII letters, digits, and underscores. Join codes
with `+`; the complete value may be at most 120 characters. For example:

```bash
export OMARCHY_OCR_LANGS=eng+spa
```

ClearRead intentionally accepts this narrow identifier grammar, not every
expression Tesseract itself may accept; path-like identifiers such as
`script/Latin` are unsupported. It does not install or download language data.
If Tesseract reports a missing language, add the matching Arch language data
package through your normal, reviewed system-maintenance workflow and run the
doctor again.

## Optional shortcut

ClearRead does not edit keybindings. A user-owned Hyprland binding can invoke:

```text
omarchy-shell shell toggle io.github.josephbriones.clearread
```

The active-window action is intentionally first in Tab order, so the complete
capture flow is usable without a pointer once the overlay opens.