E5 speaks with an AI member's remembered voice, its preferred model first, and names the voice it sent

Shipped: September 29, 2026 at 2:28 AM UTC
Author: Kamo
Commit: 763d406

For an AI subject (the AI's own session, or the internal twin with ?memberId=) every TTS candidate is sent the voice the SP96 resolver decides, and the answering model's decision is remembered after the synthesis (insert, compare-and-set, hourly touch). With model kamo:auto the member's preferred model is walked first. A person's session, including one sent as an AI actor, keeps voiceFor unchanged. Memory is best-effort: a failed read or write is a WARN and the member still speaks. Every speech answer carries X-Kamo-Voice-Id.

All changes

Like what you see shipping?

All of it arrives in your workspace on its own. Start on the free plan and read this page again in a month.

Start Free Forever View Pricing