@geekbb: Cloudflare 出品,给 AI agent 配了一台"电脑" 给 Agent 一个永不消失的工作区。SQLite 存文件,三种环境(完整容器 / shell / JS)随时开,换环境不换文件。
摘要
Cloudflare Computer is a virtual filesystem inside a Durable Object that gives AI agents a persistent workspace, using SQLite for storage and pluggable runtimes (container, shell, JavaScript). It's a preview-only open-source package.
查看缓存全文
缓存时间: 2026/08/04 18:13
Cloudflare 出品,给 AI agent 配了一台“电脑“
给 Agent 一个永不消失的工作区。SQLite 存文件,三种环境(完整容器 / shell / JS)随时开,换环境不换文件。
https://t.co/Ks9vINds8H
cloudflare/computer
Source: https://github.com/cloudflare/computer
Cloudflare Computer
Cloudflare Computer is a virtual filesystem that lives inside a
Durable Object. The Durable Object holds the authoritative state in
SQLite and exposes one pluggable execution surface through
workspace.runtime. Three backends ship today:
- Container projects the SQLite state into a sandbox container as
a real FUSE mount. A sandbox-side daemon (
computerd) mounts the state as a filesystem and syncs changes back over a capnweb RPC channel. Full Linux userland, real binaries, real network. - Isolate shell runs just-bash in a Dynamic Worker. It reaches the authoritative Workspace over Workers RPC, so there is no second store or sync round trip.
- Isolate JavaScript runs an ECMAScript module in a fresh Dynamic
Worker with structured input/results, durable relative imports,
configured libraries, Workspace-backed
node:fs/promises, and trustedws:gitandws:artifactsmodules.
A Workspace may register multiple backends under stable IDs.
workspace.runtime.exec(source, { backend }) is the single execution
entry point; the selected backend defines whether source is a shell
command or an ECMAScript module. Backends connect lazily on first use.
Workspace can also be constructed without a backend at all, giving callers the filesystem on its own.
PREVIEW ONLY This package is provided as a preview for feedback only. APIs are unstable and the design is subject to change.
Suitable for experiments, exploration and prototypes. It is NOT suitable for production use at this time.
The specification under
docs/is forward-looking — read it for intent, not as description of the code today.
Using it
If you want to build on Cloudflare Computer, install
@cloudflare/computer and follow that
package’s README — it has the installation steps, the entrypoint map,
and worked examples of the fs and runtime surfaces.
To work on this repository itself, see
CONTRIBUTING.md for setup, build, and test
instructions.
Examples
The examples/ directory holds runnable consumers of the
public surface. Each is a Worker workspace with its own README.
examples/container— runscomputerdinside a container, mounts a workspace, and talks to a Durable Object over capnweb. Awrite/read/execHTTP surface.examples/worker-shell— same HTTP surface as the container example, but the shell runs just-bash in a Dynamic Worker loaded throughenv.LOADER. No container.examples/worker-javascript— mirrorsworker-shell, butexecevaluates an ECMAScript module in a Dynamic Worker instead of running a shell command.examples/think— a@cloudflare/thinkchat agent that uses the workspace as its working directory, reachable from a terminal.examples/think-compare-runtimes— a web UI that runs the same agent task against the container and worker runtimes side by side.examples/tutorial— a step-by-step build: one endpoint, one agent that writes a markdown recipe card on the host and runspandocon it in the container to produce a PDF.examples/artifacts— generates a Worker project in a workspace and publishes it to Cloudflare Artifacts as a clone-ready repo.examples/assets— turns a prompt into an image with Workers AI, writes it to the workspace, and returns a shareable link through@cloudflare/computer/assets.
Repository layout
The repo is a small monorepo. Each package has its own README with package-specific status and usage notes.
packages/dofs(@cloudflare/dofs) — Durable Object SQLite-backed virtual filesystem, sync protocol building blocks, and a@platformatic/vfsprovider for Node.packages/rpc(@cloudflare/computer-rpc) — capnweb wire types and server/client helpers shared between the Durable Object andcomputerd.packages/computerd(@cloudflare/computerd) — thecomputerddaemon: a FUSE mount plus HTTP/WebSocket RPC server that runs inside the sandbox container.packages/computer(@cloudflare/computer) — the top-level Computer package consumed by Durable Objects. Work in progress.packages/computer-computerd-linux-x64(@cloudflare/computer-computerd-linux-x64) — the prebuiltcomputerdbinary for linux-x64, distributed for use in container images.
Performance
computerd’s FUSE mount beats real disk on metadata-heavy work and
trails it on large sequential I/O. See
docs/19_performance.md for the full fs-bench
numbers, a cloudflare/sandbox-sdk npm install comparison, and how
to reproduce them.
Documentation
docs/— design specification. Forward-looking; treat as intent.docs/19_performance.md— filesystem benchmarks.
Contributing
See CONTRIBUTING.md for setup, formatting,
testing, commit message, and pull request conventions.
If you’re working in this repo as an agent, start with
AGENTS.md and the skills under
.agents/skills/.
License
MIT. See LICENSE.
相似文章
@vikingmute: Cloudflare 除了一个刚出了 computer 的库理念非常有趣:https://github.com/cloudflare/computer… 一个在 Durable Object 里的虚拟文件系统,基于 SQLite 做持久化…
Cloudflare 发布了名为 computer 的预览库,在 Durable Object 中提供基于 SQLite 的虚拟文件系统,并支持多种运行时后端(容器、隔离 shell、隔离 JavaScript),为 Agent 提供持久化工作空间。
cloudflare/computer
Cloudflare Computer 是一个构建在 Durable Objects 上的预览版虚拟文件系统,具有可插拔的运行时,支持容器、Shell 和 JavaScript,从而在 Cloudflare Workers 上实现有状态的无服务器计算。
@geekbb: 这个好,一个开源内存层,让 Claude、ChatGPT、Cursor 和 Codex 共享一个大脑。 主要是它部署在 Cloudflare Workers 上,把记忆放在你自己 Cloudflare 账号里,对外走 MCP,数据落在用户…
Second Brain is an open-source memory layer that lets Claude, ChatGPT, Cursor, and Codex share persistent, semantic searchable memory, deployed on Cloudflare Workers for user-controlled storage.
Cloudflare OS:面向代理、应用和工作的开放平台
Cloudflare 正在开源 Cloudflare OS,这是一个开放平台,为每位员工提供基于公司上下文、技能和内部系统的 AI 代理和工作空间。
@aigclink: 面向AI智能体的一款统一虚拟文件系统:Mirage,让AI Agent用bash一招打天下,不用学N个API Mirage等于给AI搞了一个万能硬盘,各种分散的数据都被映射成同一个文件系统,AI可以用Unix命令直接操作 S3、Googl…
Mirage is a unified virtual file system for AI agents that maps disparate data sources (S3, Google Workspace, Slack, GitHub, etc.) into a single filesystem, allowing agents to use standard Unix/bash commands across all integrations without learning multiple APIs.