Three checks before your voice agent goes live, five January updates, and a few reads and listens worth your time.
Sonora Sonora
View in browser

JANUARY 2026

Sonora Shipped

2026 is off to a fast start. This issue has an editorial on running AI voice agents safely, the changes that shipped in January, and a few reads and listens worth the time.

Pre-flight: tools scoped to the agent, keys kept out of prompts, Guardrails on every turn.

THE EDITORIAL

Is that voice agent safe to run?

Voice agents are now trusted with tools, credentials, and live callers. Teams are turning them on faster than they are checking them.

The most common failures are the same three: an agent that can call any tool the workspace has, tool credentials pasted into the system prompt, and no policy check between what the model says and what the caller hears.

Before an agent goes live, run these three checks:

  1. Scope tools to the agent. Make sure a call can only invoke tools on that agent’s allow-list, not every tool in the workspace.
  2. Keep keys out of prompts. Store tool credentials as Connections and reference them by id in the agent configuration.
  3. Turn on Guardrails. Run policies as independent checks on every turn, with the ability to retry, escalate, or end the call.
Read the full article  →

Shipped in January

Guardrails audit log

Every retry, escalate, and end decision is written to the call transcript with the policy name that triggered it.

Scoped tool permissions

Tools are allow-listed per agent configuration, and a call cannot invoke a tool outside its list.

Connections

Store tool credentials once and reference them by id from any agent configuration, instead of keeping them in the prompt.

Barge-in sensitivity

Set low, medium, or high sensitivity on the Ensemble API session to control how readily the agent yields when the caller talks over it.

Recording export

Deliver call recordings to an S3-compatible bucket of your workspace’s choosing.

View the full January changelog →

Worth your time

PODCAST

Turn-taking is a product decision

Episode 12 of Sonora Signal on why interruption handling belongs to the product team, not only the audio stack.

READ

Why your voice agent needs a kill switch

How to stop an agent mid-call and what should happen to the caller when you do.

WATCH

Simulation tests, live

A recorded 20-minute session on writing simulation tests for Ensemble agents.

Until next month,
The Sonora Team

P.S. Built something on Ensemble? Share it with us for a future issue.

What did you think of this issue?

Not my thing Okay Great
You’re receiving Sonora Shipped because you opted in to updates from Sonora, the voice-agent platform behind Sonora Ensemble.
Docs  |  Blog  |  Help  |  Contact
12 Harbor Mill, Brooklyn, NY 11201
Copyright © 2026 Sonora, Inc. All rights reserved.
Privacy Policy  |  Terms
Unsubscribe  |  Email preferences