pypi · mcp-voice-summary
$uvx mcp-voice-summary==0.1.1 VOICE_ENGINE — Synthesis engine: 'edge' for Azure neural voices (needs internet), 'sapi5' for fully offline native speech.
VOICE_LANGUAGE — ISO 639-1 code such as 'es' or 'en', optionally with a region such as 'pt-BR'. 'auto' detects it per summary, which is unreliable on short text. Empty follows the system language.
VOICE_GENDER — 'female' or 'male'. Requires the edge engine; sapi5 does not expose the gender of its voices.
VOICE_RATE — Reading speed. 100 is normal, 120 is faster, '-15%' is relative.
VOICE_VOLUME — Volume, 0 to 100.
VOICE_MUTE — Set to 1 to validate and queue summaries without playing anything. Useful in shared spaces.
VOICE_MODE — 'full' speaks the sanitized summary, 'short' replaces it with a fixed generic message so no summary content reaches the speakers or the cloud engine.