8 Commits

Author	SHA1	Message	Date
jenkins	91eb4f92b7	perf(hermes): isolate rolling speech inference	2026-08-24 00:36:28 -03:00
jenkins	af5ee4697b	fix(hermes): warm STT before accepting speech	2026-08-23 23:52:30 -03:00
jenkins	d254931a14	feat(hermes): ship full-duplex voice release	2026-08-23 22:13:52 -03:00
jenkins	708d611101	feat(hermes): stream hands-free voice turns	2026-08-23 18:29:12 -03:00
Hermes Agent	820872e117	feat(hermes-voice): route Whisper language to multilingual Piper ... Port the original #27 detected-language pipeline onto the verified PR #39 prerequisite while preserving the current-main conversation instrument and host continuity changes. Keep voice selection server-side with no user selector or client voice field. Reuse 207c16ab only for its stricter exact-code trust boundary, omitting malformed or absent language so Piper defaults to Amy.	2026-08-21 13:58:50 +00:00
jenkins	5b208ae9c4	fix(hermes): isolate workers and gate noisy speech	2026-08-10 01:47:51 -03:00
jenkins	d04852cb84	perf(hermes): use conversational Whisper default	2026-08-10 01:30:50 -03:00
jenkins	4d029f85b7	feat(hermes): add private voice and isolated workflows	2026-08-10 00:43:10 -03:00