Follow Discord
Sweep 22 Oct 2026 · 04:32Z Build v2.1.308 527 read Stable v2.1.308 Latest v2.1.308 Next v2.1.308 Feeds RSS JSON llms.txt llms-full.txt Unofficial

Claude Code v2.1.304 ·

Windows sandbox works out its file access from credentials and allowed domains

On Windows, the sandbox now uses its credentials and allowed network domains when working out which files it may reach

You'll notice Improvements
JSON All of v2.1.304
You'll noticeTier: how much it should matter to you
2Useful: my rating, 1 to 5
2Signal: worth watching, 1 to 5
SandboxArea: what it touches
ImprovementsKind: in v2.1.304,
ImprovementsSection of the release

Unclear Which files the Windows sandbox can now read or write as a result is not stated.

What

The sandbox is a set of limits on what commands run by Claude can reach. On Windows, Claude Code sets it up by working out which files and folders the sandbox is allowed to use. That calculation now also takes the sandbox's credentials and its list of allowed network domains (allowedDomains) into account. Before, it was worked out from the sandbox configuration alone.

When TLS termination is turned on, setup also runs an extra first step. TLS termination means the sandbox opens up encrypted web traffic so it can be checked.

Why

This could change which files the Windows sandbox is able to read.

Read from
Names in the bundleallowedDomains
Since it was published

The entry above is what we published on the day. These lines were added later, as Anthropic's own pages caught up, and they sit beside the original rather than replacing it.

Added since A small documentation edit on session-operations touched a line naming allowedDomains after this was published. You can update a session's `agent.tools` (csharp, go: `Agent.Tools`) and `agent.mcp_servers` (java, php: `agent.mcpServers`; csharp: `Agent.McpServers`; go: `Agent.MCPServers`), including permission policies and per-tool web settings such … managed-agents/session-operations see the edit
Confirmed since Anthropic's documentation has since written up allowedDomains, on secure-deployment. Set `allowed_domains` (java, php: `allowedDomains`; csharp, go: `AllowedDomains`) on `web_fetch` to the sites the agent needs. Any URL the agent fetches can carry data to the site it points to, so this list is a security control. On `web_s… managed-agents/secure-deployment see the edit
How sure we are
Something disagreesSomething we can check disagrees with this entry, or the writer said they could not settle it.
The writer flagged doubtWhich files the Windows sandbox can now read or write as a result is not stated.
Anthropic's documentation agreesAnthropic's documentation has since written up allowedDomains, on secure-deployment.

See this entry in the whole of v2.1.304 →

Feedback