clauder: Command for Claude Code

Toggle spoken summaries

Turn the spoken summary of each response on or off. The setting is audio_summary.enabled in .claude/preferences.json. With no argument, toggle it. With on or off, set it to that state.

Steps

1. Read .claude/preferences.json. Treat a missing audio_summary key as disabled.
2. Decide the new state from the argument (on, off, or none for toggle).
3. Write the new state with this single command, replacing STATE with true or false. Do not use the Edit or Write tools on this file; it is protected by the immutable-patterns hook and this command is the supported way to change it:

python3 - STATE <<'PY'
import json, sys
path = ".claude/preferences.json"
try:
    data = json.load(open(path))
except (FileNotFoundError, json.JSONDecodeError):
    data = {}
data.setdefault("audio_summary", {})["enabled"] = sys.argv[1] == "true"
json.dump(data, open(path, "w"), indent=2)
open(path, "a").write("\n")
print("audio_summary.enabled =", data["audio_summary"]["enabled"])
PY

1. Reply in one line with the new state, for example "Spoken summaries are now on." If turning on, add that say on macOS or espeak-ng/espeak on Linux must be installed, and that the first spoken summary follows this response. Do not speak this confirmation yourself; the Stop hook handles speaking.

Only the audio_summary key changes. Leave every other preference as it is.