Claude Code has received updates that include new features, improvements, and bug fixes. These changes enhance control over tool execution, introduce caching mechanisms, and provide more flexibility for self-hosted environments.
Key Points
- A
--restrictedflag (orCLAUDE_CODE_RESTRICTED=1) has been added to remove built-in tools that run commands or code and WebFetch, unless specifically named in--tools. - The restricted mode keeps file tools within the working directory, refuses
bypassPermissions, and ignores user, project, and local settings files. - An
experimental.cacheTtlsetting (e.g.,"5m"or"1h") is now available in agent frontmatter for a per-agent prompt cache TTL, used when no subagent TTL is configured. claude self-hosted-runner --client-label <label>(orSELF_HOSTED_RUNNER_CLIENT_LABEL) allows overriding the label a runner registers with, defaulting to the hostname.- Server-managed settings diagnostics now include a startup warning for settings load failures and
/doctorand/statuslines explaining load issues or why settings were not fetched. - A warning in
/web-setupindicates when the GitHub CLI token lacks the workflow scope. /usage-creditsis available for Enterprise organizations billed via AWS Marketplace, self-serve Enterprise, and Enterprise trials, allowing members to request higher usage limits from administrators.
Context
According to the Claude Code changelog, these updates aim to provide users with more granular control over the environment and improve diagnostic capabilities. The documentation index can be fetched at /docs/llms.txt to discover all available pages before further exploration.
Why It Matters
These updates offer builders enhanced security controls through the restricted mode, improved performance via prompt caching, and better manageability for self-hosted deployments. The new diagnostic and usage credit features also streamline operational aspects for enterprise users.
What To Do
- Review the
--restrictedflag documentation to understand its implications for tool usage and permissions. - Experiment with the
experimental.cacheTtlsetting in agent frontmatter to optimize prompt caching for specific agents. - Note the new
claude self-hosted-runner --client-labeloption for custom labeling of self-hosted runners. - Check the
/doctorand/statusendpoints for server-managed settings diagnostics if encountering load failures.
