A custom status line for Claude Code that
shows the model name and current effort level, percentage of the context
window consumed (color-coded green/yellow/red), exact token counts, whether
the prompt cache is warm, the remaining rate-limit percentages with time until each window resets
(whichever of the 5-hour and weekly windows your plan has), the API-equivalent
session cost, a second line with the session name, PR number and session ID,
plus the familiar user@host:path (branch) prefix.
Voice-to-text transcription using OpenAI Whisper, with integrations for AI coding assistants and Emacs.
- openai-whisper:
pip install openai-whisper - sox:
brew install sox(for thereccommand) - ffmpeg:
brew install ffmpeg