Claude is becoming more than a chatbot for answering questions. Anthropic is bringing document creation, presentation tools, and larger workspaces into the same interface, while Coder is adding Claude Code support for companies that need tighter control over where software agents run.
The changes point to one main idea: users should not have to jump between separate tools for every stage of a project. Claude can now create editable documents and slides inside conversations, and teams can use Claude Code through Coder’s infrastructure when their work requires security, network controls, and audit records.
Claude brings documents and presentations into chat
Anthropic has launched Claude Docs, a feature that lets users ask Claude to create documents. A request for a one-page sheet with pertinent information, for example, produces the document inside the chat instead of leaving the user with plain text alone.
The document starts as a private workspace. Users can then choose to share it with colleagues, edit it inside Claude, or export it and open it in Google Docs or Microsoft Word. That gives the feature a practical role in everyday work, from preparing information sheets to refining material before a wider team sees it.
Anthropic is also adding slide and presentation creation to conversations. Claude can drop slides into the chat, where users can edit them or download them as PowerPoint or PDF files. Documents and presentations now sit closer to the point where people ask questions, gather information, and shape their first draft.
These tools are part of Anthropic’s effort to bring Claude’s separate work features together. The company is merging Cowork and Artifacts into a single interface under the name “One Claude.”
One Claude connects different kinds of work
Cowork is a mode inside the Claude app that uses agentic AIs to complete multi-step tasks. Artifacts are pieces of content, including documents, HTML websites, and diagrams, that Claude saves in a sidebar for easy access.
Design is a visual workspace, while Cowork is built for bigger work. Before this change, those spaces did not carry tasks from one to the other. The new setup is meant to make the capabilities available from any conversation, with the context, skills, and connectors already in use.
Anthropic’s Cat Wu, head of product for Claude Code, is part of the company’s broader push to make Claude handle more of the work around a request. Instead of treating writing, visual projects, and multi-step tasks as separate destinations, the interface brings them into one place.
The updates are rolling out to Pro and Max subscribers over the next few weeks on web, desktop, and mobile. Users on Team and Free plans will also get the features “soon.”
Claude Code gets a controlled home for enterprise work
On September 15, 2026, Coder announced Claude Code support through its new Agent Relay. Agent Relay is a self-hosted execution environment for cloud coding agents that launched on September 2, with Cursor as its first agent provider.
With the new support, Claude Code can run inside Coder workspaces on a customer’s own infrastructure. Those workspaces can use network-governed, sandboxed, and fully auditable environments, which gives companies control over the machines and systems touched by the agent.
Anthropic still handles billing and runs the agent loop, but everything Claude Code touches lives on machines controlled by the customer. Each workspace starts a Claude Code runner that connects to Anthropic’s backend, and every run creates an audit record covering workspace provisioning and its lifecycle.
Teams can also add process-level network policies that block and log requests. Turning on Claude Code in existing platform setups requires no new deployment or security review, which removes a step for organizations that already use those setups.
Claude Code support in Coder Agent Relay is in early access with select design partners. Coder describes itself as the AI Operating Layer for the enterprise, providing self-hosted, model- and cloud-agnostic infrastructure for AI. The company says more than 125,000 GitHub stars point to the reach of its work.
Josh Epstein, president at Coder, said, “Claude Code is one of the agentic tools our enterprise customers ask for by name, and regulated industries have wanted it the most. Agent Relay removes the barrier, giving organizations the same Claude Code experience their teams want while executing within infrastructure they control.”
Documentation and deployment guides are available from Coder for the Agent Relay setup. Together, the Claude interface updates and the Coder integration show Anthropic pushing Claude in two directions at once: easier creation for everyday users and controlled execution for regulated industries.
Based on
- Anthropic’s new Claude Code feature could drain your plan before lunch — thenewstack.io
- Anthropic’s Claude Can Now Create Editable Documents For You — engadget.com
- Coder Brings Claude Code to Agent Relay, Unlocking Agentic Development for the World’s Most Regulated Enterprises | Markets Insider — markets.businessinsider.com
