Claude Code's 'Chief of Staff' Update Ties Multiple Coding Tasks Together

Anthropic redesigned Claude Code's Projects feature to let developers manage multiple related tasks within a single project, carrying context across assignments instead of requiring separate conversations. For lean engineering teams, this means less time re-explaining goals and fewer conflicting changes, though trust in AI-generated code remains a real concern.

A Projects update that ties multiple coding tasks together

Before this update, a developer working on a login bug, a payment system update, and a batch of tests had to split the work across separate Claude conversations, each needing the same background context. The redesigned Projects feature lets you hand off all three tasks within a single Claude Code project. Claude carries the context across assignments and adjusts its workflow as the project changes, according to the announcement.

Anthropic describes the effect as moving Claude from a chatbot that handles one request at a time toward a code-writing chief of staff that coordinates interdependent work and tracks how each change affects the others. Instead of fixing the login bug without knowing the payment system was also being updated, Claude can adjust both together.

Why this matters for lean engineering teams

Small teams that ship quickly stand to benefit most. Y Combinator CEO Garry Tan noted in March 2025 that AI generated 95 percent of the code at about a quarter of startups in its winter cohort, and that some 10-person teams were producing as much work as teams of 50 to 100 people. Coordination overhead has been a bottleneck, and the new Projects feature directly addresses that by keeping related tasks connected.

But generating more code with AI creates its own friction. A 2026 survey of over 1,100 developers found that while AI helped write 42 percent of their code, 96 percent did not fully trust the results, and 38 percent said reviewing AI-generated code took more time than reviewing code written by a person. By maintaining context across tasks, the Projects update could reduce the chance of conflicting changes and the associated review time, but it won't eliminate the need for careful human review.

What changes for builders using Claude Code

The updated Projects feature is initially available in beta to select Claude Pro and Max customers. Anthropic plans to expand access to more users over the coming week and eventually bring the feature to Team and Enterprise plans. If you're on Pro or Max, you can try the beta now. For teams on Team or Enterprise, the timing is still unconfirmed.

The limits you should still watch for

Better coordination reduces some risk, but AI-generated code still carries trust and accuracy concerns. The survey data makes clear that developers spend significant time validating AI output. The feature is also in beta, so edge cases and unexpected conflicts may still surface. The ability to track cross-task impact depends on how well Claude understands the project's full scope, which may vary with project complexity. Competing tools like OpenAI's Codex and Amazon's Kiro are also pushing in similar directions, so this space will keep evolving.