add-audio-format

name	add-audio-format
description	Add support for a new audio file extension (e.g. .aiff, .wma) to AudioToText, wiring the file-picker filter and deciding native-vs-ffmpeg-conversion. Use when the user wants the app to accept another audio format.
allowed-tools	Bash, Read, Edit

add-audio-format — support a new audio extension

Add an audio extension end-to-end so users can pick it and transcribe it.

Decide: native or converted?

whisper-cli decodes only wav, mp3, ogg, flac directly. Everything else must be converted to 16 kHz mono WAV via ffmpeg first.

• If the new format is one whisper-cli reads natively → add it to nativeFormats in transcribe.go.
• Otherwise → leave it out of nativeFormats; the existing convertToWav path in transcribe.go handles it automatically (requires ffmpeg).

When unsure whether a format is native, treat it as needing conversion — the fallback path is safe.

Steps

1. File picker filter — add the extension (with leading dot, lowercase) to the storage.NewExtensionFileFilter list in chooseFile in main.go.

2. Native decode (only if native) — add the extension to the nativeFormats map in transcribe.go.

3. Docs — update the supported-formats list in README.md, placing the extension under "directly" or "via auto-conversion" accordingly.

4. Verify — run the project gate:

   go build ./... && go vet ./... && gofmt -l .
   

   (Or invoke the check skill.)

Notes

• Extensions are matched case-insensitively (strings.ToLower / strings.EqualFold), but list them lowercase with a leading dot.
• Do not duplicate an extension that is already present.
• Conversion always produces a temp WAV that is removed via defer os.Remove; do not change that cleanup behavior.