Jellypod, Inc.

The Claude Code Changelog

TechnologyNews

Listen

All Episodes

Claude Code Auto Mode Hits Enterprise Clouds

We break down Claude Code 2.1.158’s new Auto Mode support for Bedrock, Vertex AI, and Palantir Foundry, and what it means for developers working inside tightly controlled enterprise environments.

Plus: the new OpenTelemetry audit flag, full autonomous write-run-correct loops, and the fix for stubborn Git worktree locks.

This show was created with Jellypod, the AI Podcast Studio. Create your own podcast with Jellypod today.

Is this your podcast and want to remove this banner? Click here.


Chapter 1

Auto Mode Breaks Cloud Provider Barriers

Lachlan Reed

Welcome to the show, everyone! I'm Lachlan Reed, here with James Turner, and James, if you've ever had your workflow ground to a halt by enterprise security red tape, you'll appreciate this. [excited] Claude Code version 2.1.158 just dropped, and it finally brings Auto Mode to Bedrock, Vertex AI, and Palantir Foundry for Opus 4.7 and 4.8.

James Turner

Oh, that is massive! [excited] So developers stuck behind those corporate cloud gateways because of strict data compliance don't have to suffer through "approval fatigue" anymore. They can actually use the agent's full autonomous loop.

Lachlan Reed

Exactly. No more babysitting the terminal. To kick this off, you just set `CLAUDE_CODE_ENABLE_AUTO_MODE=1` in your environment. Once that's live, the agent can execute complete write-run-correct cycles entirely on its own—no manual keypress approvals needed for every single file change or shell execution.

James Turner

Right, but giving an agent free rein to execute shell commands and modify code autonomously inside an enterprise VPC... [skeptical] that's going to make security teams sweat. How are they handling the paper trail?

Lachlan Reed

They actually thought of that. They've introduced a new OpenTelemetry flag: `OTEL_LOG_TOOL_DETAILS=1`. When you enable that, it captures the exact tool parameters and inputs the agent uses. It's a complete audit trail of what the agent tried to do and when.

James Turner

Okay, so security gets their deep telemetry, and developers get their speed. [thoughtfully] That's a fair trade. And hey, did they fix that annoying Git lock issue?

Lachlan Reed

They did! [chuckles] They finally fixed the bug that was leaving git worktrees locked. Now, when the agent is done, your workspace cleanup is actually clean. No more manual unlocking faff.

James Turner

Thank goodness. This really feels like Claude Code is maturing from a cool local CLI toy into something you can actually run at scale in a highly regulated environment. Powered by Jellypod AI.

Lachlan Reed

Too right, mate. Let's see how enterprise teams adopt it. See you all next time!