Anthropic shipped a meaningful update to Claude Code this week. The new features target teams using Claude Code in production environments: shared configuration, team-level analytics, and centralized policy management.
For organizations adopting Claude Code at scale, the update closes governance gaps that made enterprise use awkward.
The new features
Team profiles. A team can publish a shared CLAUDE.md file (or a directory of profile files) to a team registry. Team members’ Claude Code sessions automatically incorporate the team’s standard configuration. Updates propagate without each developer manually pulling.
Usage analytics. Team admins can see usage by member, broken down by project, tool invocations, and cost. The dashboard is similar to what Cursor 1.0 introduced but specifically scoped to Claude Code’s CLI flows.
Policy enforcement. Teams can set policies that apply to all members: which models are allowed, which tools (filesystem write, shell execution) require additional confirmation, what data types are excluded from prompts. Policies are enforced client-side.
Shared MCP server registry. Teams can publish vetted MCP servers to a team registry. Members install from the team registry; servers go through team review before being available.
Why this matters
Claude Code has been popular among individual users since launch. Team adoption has been more cautious because:
- Configuration was per-user; ensuring consistency required out-of-band coordination
- No usage visibility; managers couldn’t see if the tool was being used effectively
- No policy controls; preventing risky usage patterns required documentation rather than enforcement
This update addresses each gap. Organizations that had been hesitant to deploy Claude Code at scale now have the governance features to do so.
A specific use case
For a team standardizing on Claude Code:
- The team lead writes a comprehensive CLAUDE.md with the team’s coding standards, architectural conventions, testing requirements, etc.
- They publish it to the team registry.
- New team members get the standardized configuration on first run.
- Updates to the standard propagate automatically.
Compared to “tell new members to set up Claude Code and copy this CLAUDE.md from the wiki,” this is a meaningful improvement in operational reliability.
How this compares to the editor agents
Cursor and Cline have similar features for editor-based AI tools. The Claude Code update brings the CLI category to feature parity on team management.
For organizations that are CLI-heavy (devops, backend, certain data engineering workflows), Claude Code’s terminal-first approach matches the workflow better than editor-based tools. The team features remove the gap that previously favored editor tools for organizational adoption.
What’s missing
A few things I’d want that aren’t in this release:
Per-project profiles. The team profile is global per team. Different projects within a team often have different conventions. Per-project profile inheritance would help.
Plain-text policy export. The policies are configured via Anthropic’s UI. Some teams want to manage policies via Git (for review, history, versioning). A plain-text policy file format would address this.
Audit trail of policy changes. Who changed what policy when. The current dashboard shows usage but not configuration history.
These are reasonable next steps. The current release is enough to start; the gaps are non-blocking for most teams.
The competitive view
Claude Code competes with:
- Codex CLI (OpenAI): newer, similar shape, less mature on team features
- Aider: open source, no team features
- Continue CLI: smaller user base
For team adoption specifically, Claude Code is now meaningfully ahead. Codex CLI doesn’t have an equivalent team feature set yet. Aider, by design, is single-user. Continue CLI hasn’t prioritized teams.
If your organization is choosing a CLI agent for standardized adoption, Claude Code is the obvious choice as of this release.
Pricing implications
The team features are bundled with Anthropic’s existing team plans. No separate Claude Code subscription. Organizations paying for Claude Pro/Max get the team features included.
This is more generous than Cursor’s pricing, where some team features require Business or Enterprise tiers. Anthropic seems to be optimizing for adoption rather than per-feature monetization.
Worth adopting?
For teams already using Claude Code informally: yes. The team features add real value at no extra cost.
For teams using other CLI agents but considering Claude Code: this update may be the trigger to evaluate. If the team management story matters to your organization, Claude Code’s just-shipped features are competitive.
For teams using only editor-based AI tools: the question is whether your work has meaningful CLI-shaped components. If yes, Claude Code is worth adding alongside your editor tool. If no, this release doesn’t change much.
The CLI agent category
This release confirms the CLI agent category is mature enough for team adoption. The features that matter for organizations — shared config, analytics, policy enforcement — now exist in the leading product.
Six months ago, “we use Claude Code as a team” required out-of-band coordination. Now it requires clicking some setup buttons. The friction reduction is real.
For organizations evaluating AI tooling broadly, the CLI agent category is now part of the consideration set, not a curiosity. The shape of “developer at terminal asks AI to do things” is a workflow with real organizational support.