Claude Platform Makes Computer Use, Browser Use, Skills, and Files Generally Available
Anthropic says computer use, browser use, the Skills API, and the Files API are now generally available on the Claude Platform, combining multi-action software control with versioned skills and reusable files for production agents.
Anthropic says four Claude Platform building blocks reached general availability on August 20: computer use, a new browser use tool, the Skills API, and the Files API. The launch gives developers one official path for agents that operate software, apply reusable procedures, and read or write files. (ClaudeDevs on X; Anthropic launch blog)
The pieces are designed to work together rather than as isolated endpoints. An agent can read an intake document, follow a versioned skill, complete a workflow in a web application, and return a finished file. (Anthropic launch blog)
Computer use takes multiple actions per turn
The updated computer use tool lets Claude operate visible software by clicking, typing, and scrolling. Anthropic says Claude can now take several actions per turn instead of one action per model call, which can reduce the number of calls needed to finish a task. (ClaudeDevs on X; Anthropic launch blog)
That is a workflow change for teams building desktop or application automation: the agent can make a sequence of UI moves before returning control to the model. Anthropic reports one customer workflow improving from 32 minutes to 13 minutes and says cost per task fell about 30% across the workflows tested, but those are reported customer results rather than a guarantee for every application. (Anthropic launch blog)
Browser use adds page structure to web automation
The new browser use tool extends computer use to web applications. Alongside a screenshot, it reads the page structure so an agent can target fields and buttons by their elements rather than relying only on screen positions. (ClaudeDevs on X; Anthropic launch blog)
That distinction matters when a web page shifts its layout. The browser tool is intended to make web actions less dependent on fixed pixels, while still using the same multi-action turns as the updated computer use tool. Anthropic says the computer use and browser use tools are coming soon to Google Cloud's Vertex AI; existing beta integrations continue to work during migration. (Anthropic launch blog)
Skills API packages team procedures
The Skills API gives teams a simpler way to upload and version reusable skills. Anthropic describes a skill as a package of instructions, scripts, and templates that Claude loads when a task calls for it; skills run in Claude's code execution sandbox, so the team does not need to host the skill itself. (ClaudeDevs on X; Anthropic Skills documentation; Anthropic launch blog)
A team can pin a request to a skill version or use the latest version. That makes the procedure an explicit dependency of the request instead of an undocumented prompt fragment. The Skills API is also available through Microsoft Foundry, according to Anthropic's launch post. (Anthropic launch blog)
Files API keeps reusable documents available
The Files API lets a team upload a document once and reference it by file_id in later requests. Agents can read documents such as PDFs or spreadsheets and download files they create, avoiding a new upload for every request. (ClaudeDevs on X; Anthropic launch blog)
The August 20 launch also adds automatic file expiration, 5x higher rate limits, and 1 TB of storage per organization. Anthropic does not state a new price in the launch blog, so teams should check their account and current platform terms before estimating cost. (Anthropic launch blog)
What developers should take from the release
The practical sequence is straightforward:
- Store the input document with the Files API.
- Attach a versioned skill that describes the team's procedure.
- Use browser use or computer use to complete the task in the target application.
- Save the resulting document or confirmation as a file.
This is a meaningful platform update for production agents because it joins software control, reusable expertise, and document handling in one generally available release. It does not establish universal availability across every cloud provider or promise the reported customer performance for every workflow. (Anthropic launch blog)
Sources
- ClaudeDevs on X, "Computer use, the browser tool, the Skills API, and the Files API are now generally available": https://x.com/ClaudeDevs/status/2090540270219567575
- ClaudeDevs on X, "Claude now takes several actions per turn": https://x.com/ClaudeDevs/status/2090540271557546261
- ClaudeDevs on X, "browser_toolset_20260801": https://x.com/ClaudeDevs/status/2090540272690057481
- ClaudeDevs on X, "The Skills API and Files API": https://x.com/ClaudeDevs/status/2090540273939996958
- ClaudeDevs on X, "The Files API lets you upload once": https://x.com/ClaudeDevs/status/2090540275357606263
- Anthropic, "Build production agents with computer use, the Skills API, and the Files API": https://claude.com/blog/computer-use-skills-api-files-api
- Anthropic, "Agent Skills": https://platform.claude.com/docs/en/agents-and-tools/agent-skills/overview
Read next
Keep building the workspace playbookClaude Platform's 'ant' CLI Brings the Full Claude API to Your Terminal
Anthropic's Claude Platform documents 'ant', a command-line tool that exposes every Claude API resource as a subcommand. It sends Messages requests, browses responses, version-controls agents and environments, and runs a self-hosted Managed Agents worker, with Claude Code able to drive it natively.
Claude Managed Agents Add Self-Hosted Sandboxes (Public Beta) and MCP Tunnels (Research Preview)
Anthropic says Claude Managed Agents can now run tool execution in a sandbox you control (public beta) and connect to private MCP servers via MCP tunnels (research preview). The update targets enterprise security requirements by keeping execution and private services within an organization's perimeter.
Claude Managed Agents Add Memory Stores, Domain Controls, and a Redesigned Console
Anthropic's Claude Managed Agents updates let self-hosted sandbox sessions attach memory stores, restrict web_search and web_fetch with allowed_domains or blocked_domains, and inspect multi-agent sessions in a redesigned Claude Console viewer.
Frequently Asked Questions
What is generally available in Anthropic's August 20 release?
Anthropic says computer use, the browser use tool, the Skills API, and the Files API are generally available on the Claude Platform. The browser use tool is a new addition to computer use for work in web applications.
How does Claude's updated computer use tool work?
Claude can take several actions per turn instead of one action per model call. Anthropic says this can let tasks finish in fewer calls and less time; the blog does not promise a fixed improvement for every workflow.
What do the Skills API and Files API add?
The Skills API lets teams upload and version skills made from instructions, scripts, and templates. The Files API lets agents reuse uploaded documents by file ID and download files they create; the launch includes automatic expiration, higher rate limits, and 1 TB of storage per organization.