AI News
商业产业 Hacker News Show HN AI 2026-08-27T13:02 正文 ok 1059 words

Show HN：Derive——AI 工件与工作流的开放归宿

英文标题：🔶 Show HN: Derive – An open home for AI artifacts and workflows

Derive 是一个面向 AI Agent 产出物的工作空间，可为 HTML、Markdown、演示文稿和完整网站提供持久 URL、版本历史、评论与审核能力。开发者可通过浏览器、CLI、HTTP API 或 MCP 发布内容；匿名草稿无需账号或安装即可生成实时链接和认领链接。该项目支持将 Agent 的指令、技能、数据源、权限与执行记录沉淀为可复用资源，并让 Agent 基于评论发布后续修订。Derive 采用 FSL-1.1-ALv2 许可证，可通过单个 Node 容器自托管，支持 SQLite/本地磁盘或 Postgres 与 S3/R2 等存储组合。
打开原文相关 2 · 机会 1
media38.4 KB media39.5 KB media1.3 KB media1.3 KB media1.3 KB media591.7 KB media134.1 KB media145.9 KB media6.0 KB media58.5 KB

中文

Derive

发布、审核并拥有你的 AI 工件。

为每个版本、评论、修订和决策提供一个持久 URL。你的团队审核工作；任何兼容的 Agent 都可以根据反馈采取行动。可使用托管服务，或通过一个自托管容器运行。

免费试用 · 自托管 · 文档

许可证：FSL-1.1-ALv2 npm @derive-to/cli npm @derive-to/mcp

Derive 资料库：带实时预览的已发布工件

无需账号或安装。一次请求即可发布实时页面：

curl -F file=@page.html https://derive.to/v1/drafts   # or a .zip of a whole site
# → a live URL, plus a claim link that turns the draft into a
#   permanent, versioned artifact in your workspace

为你的 Agent 提供一个可持久发布、查找和改进工作的空间：

npx skills add derive-to/derive --skill derive        # any agent that reads skills
claude mcp add --transport http derive https://derive.to/mcp   # or connect over MCP

或者将以下内容粘贴到兼容的编程 Agent 中，让它自行完成设置：

I'd like you to set up Derive: where the work we make gets published, reviewed, and kept.

Install the skill if I have npm: npx skills add derive-to/derive --skill derive
Otherwise, read https://derive.to/skill.md and follow it (MCP server: https://derive.to/mcp).

Then pick a plan, report, or designed page we've worked on recently and publish it. If I'm not connected to Derive, use the anonymous draft flow. It needs no account and returns a live URL with a claim link. Send me both links.

Derive 是一个用于管理由 Agent 创建工作的工作空间。你可以为 HTML 页面、Markdown 文档、演示文稿或已构建的网站赋予持久 URL 和版本历史。可从浏览器、CLI、HTTP API，或通过 MCP 连接的兼容 Agent 发布内容。你可以将工作设为私有或分享出去，留下评论，直接编辑，并在同一 URL 发布下一个版本。

内容、版本历史和评论都会随各个工件保存。队友或另一个 Agent 可以打开同一 URL，查看发生了什么变化，无需从聊天记录中重新还原工作内容。

一个工作空间还可以保存 Agents：具有指令、技能、数据源、权限和执行连接的可复用工作者。它们可以回答问题、承担工作，并将结果发布到同一个工件资料库中。

Derive 采用 Fair Source 模式，并支持自托管。你可以在自己的基础设施上以一个容器运行它，也可以使用托管应用。

在其永久 URL 上渲染的已发布工件 |

发布任何内容，获得永久 URL

HTML、Markdown，或整个已构建网站。每次修订都会在同一 URL 创建一个新版本，因此你上周分享的链接仍然有效，并且仍能显示其历史记录。
---|---

评论和变更与工作内容一同保存

分享一个工件，并 @提及应参与评议的人员或 Agent。评论会锚定到确切文本，直接编辑会创建新版本，已连接的 Agent 可以读取反馈并发布针对性的修订。当某项决策确实需要签核时，可请求对该版本进行正式审核。 已打开评论和审核面板的工件
Derive 资料库：每个已发布工件均带实时预览

已发布工作的资料库

每个工件都会进入一个带实时预览的资料库。你可以置顶常用工作，找到他人分享给你的内容，并在集合中整理相关工件。
---|---

还包括：

  • 可移植的工作。 内容、版本和评论会随工件保存，因此其他人或 Agent 可以继续这项工作。
  • Agents。 在不同问题和委派工作之间复用相同的指令、技能、数据源、权限和运行历史。
  • 检查点。 将进行中工作的状态保存到一个页面中，供后续会话打开。
  • 沙箱化查看器。 工件运行在不透明源上，与 Cookie 及其他工件隔离。
  • 灵活存储。 可使用 SQLite 和本地磁盘，也可在规模化场景中使用 Postgres 与 S3/R2。
  • 实时协作。 评论、可选审批和在线状态通过 Server-Sent Events 更新。
  • 分享预览。 链接可在 Slack、Discord、X 和 Notion 中显示工件预览。
  • CLI 和 MCP。 可从终端发布，或连接兼容的 Agent。
  • 访问控制。 可将工作设为私有、对工作空间可见或公开。公开链接还可设置密码。

当前路线图位于 derive.to/roadmap，其历史记录也保存在同一 URL。

最快路径，无需安装。

  1. 前往 derive.to
  2. 创建账号
  3. 发布你的第一个工件
你将获得资料库、浏览器内发布、版本历史和评论功能。 一个容器即是完整产品：API、Web、登录、发布、评论和沙箱化查看器，以及存储于一个卷中的 SQLite 和 blob。请按照 自托管快速入门 安装使用摘要固定的发布版本，或构建当前检出版本。它包含安全的首位用户初始化、就绪检查和经过验证的首次备份。有关 Postgres、S3/R2 和云主机，请参阅部署指南。

想要一个完整工件，而不是空白起始模板？可打开或发布 官方示例 之一：一个精心设计的发布页面、一份研究简报或一份持续更新的状态报告。每个示例展示不同的起点。

npm i -g @derive-to/cli
derive init my-doc --template slides   # templates: md · html · slides
cd my-doc
derive publish                         # versioned URL; the id is saved to derive.json

连接你的编程 Agent（MCP）

添加到 Cursor 或为 Claude Code 添加一行命令：claude mcp add --transport http derive https://derive.to/mcp

# Derive is a remote MCP server (OAuth). Connect either client:
claude mcp add --transport http --scope project derive https://derive.to/mcp
codex mcp add derive --url https://derive.to/mcp

# or run a local stdio server (set DERIVE_SERVER; DERIVE_TOKEN for a static bearer):
npx -y @derive-to/mcp

Agent 将以你授予的角色权限执行操作。拥有发布权限时，Agent 可直接发布，并将其作为已保留、可恢复的版本，同时通知工作空间；较低权限范围则会以评论形式提出修改建议，供人工应用。完整设置和工具指引位于 packages/mcp/SKILL.md。

一个 Node 容器即是完整产品；存储可通过接口灵活替换。同一镜像可使用 SQLite 和本地磁盘自托管，可使用 Postgres 和 S3/R2 扩展，也可运行在 Cloudflare Workers 上。

apps/api          HTTP API, sandboxed artifact serving, viewer
apps/web          web UI (TanStack Start, SPA mode, static bundle)
packages/core     domain: ports, publish, markdown render, viewer shell
packages/db       MetaStore: sqlite (default) · postgres · d1
packages/storage  BlobStore: fs (default) · s3/r2
packages/cli      derive init (md/html/slides) · derive publish <file|dir> · derive runner serve (run an Agent)
packages/mcp      Local compatibility MCP: eight agent tools + derive://guide

每个工件均包含 OG 和 Twitter 元数据以及一个 oEmbed 文档，提供实时 Server-Sent Events 流，并在不透明源上的严格沙箱 CSP 下渲染。有关创作和嵌入细节，请参阅工件创作标准。

Functional Source License（FSL-1.1-ALv2），采用公平源代码模式。除将 Derive 作为竞争性商业产品或服务提供外，你可以为任何目的自由运行、修改和自托管 Derive。每个发布版本会在发布两年后自动转换为 Apache-2.0 许可证。详见通俗易懂的许可指南。

English / 原文

Derive

Publish, review, and own your AI artifacts.

One durable URL for every version, comment, revision, and decision. Your team reviews the work; any compatible agent can act on the feedback. Hosted, or one self-hosted container.

Try free · Self-host · Docs

License: FSL-1.1-ALv2 npm @derive-to/cli npm @derive-to/mcp

The Derive library: published artifacts with live previews

No account or install is needed. One request publishes a live page:

curl -F file=@page.html https://derive.to/v1/drafts   # or a .zip of a whole site
# → a live URL, plus a claim link that turns the draft into a
#   permanent, versioned artifact in your workspace

Give your agent a durable place to publish, find, and improve work:

npx skills add derive-to/derive --skill derive        # any agent that reads skills
claude mcp add --transport http derive https://derive.to/mcp   # or connect over MCP

Or paste this into a compatible coding agent and let it set itself up:

I'd like you to set up Derive: where the work we make gets published, reviewed, and kept.

Install the skill if I have npm: npx skills add derive-to/derive --skill derive
Otherwise, read https://derive.to/skill.md and follow it (MCP server: https://derive.to/mcp).

Then pick a plan, report, or designed page we've worked on recently and publish it. If I'm not connected to Derive, use the anonymous draft flow. It needs no account and returns a live URL with a claim link. Send me both links.

Derive is a workspace for work made with agents. Give an HTML page, Markdown document, deck, or built site a lasting URL and version history. Publish from the browser, CLI, HTTP API, or a compatible agent over MCP. Keep the work private or share it, leave comments, edit it directly, and publish the next version at the same URL.

The content, version history, and comments stay with each artifact. A teammate or another agent can open the same URL and see what changed without reconstructing the work from a chat.

A workspace can also keep Agents : reusable workers with instructions, skills, sources, permissions, and an execution connection. They answer questions, take on work, and publish results to the same artifact library.

Derive is Fair Source and self-hostable. Run it as one container on your own infrastructure or use the hosted app.

A published artifact rendered at its permanent URL |

Publish anything, get a permanent URL

HTML, Markdown, or a whole built site. Every revision is a new version at the same URL, so a link you shared last week still resolves, and still shows its history.
---|---

Comments and changes stay with the work

Share an artifact and @mention the people or agents who should weigh in. Comments pin to the exact text, direct edits create new versions, and connected agents can read feedback and publish focused revisions. When a decision truly needs sign-off, request a formal review of that version. An artifact with its comment and review panel open
The Derive library: every published artifact with a live preview

A library for published work

Every artifact lands in a library with a live preview. Pin the work you use often, find work shared with you, and organize related artifacts in collections.
---|---

Also included:

  • Portable work. Content, versions, and comments stay with the artifact so another person or agent can continue the work.
  • Agents. Reuse the same instructions, skills, sources, permissions, and run history across questions and delegated work.
  • Checkpoints. Save the state of ongoing work in a page that a later session can open.
  • Sandboxed viewer. Artifacts run on an opaque origin, isolated from cookies and other artifacts.
  • Flexible storage. Use SQLite and local disk, or Postgres and S3/R2 at scale.
  • Live collaboration. Comments, optional approvals, and presence update over Server-Sent Events.
  • Share previews. Links show an artifact preview in Slack, Discord, X, and Notion.
  • CLI and MCP. Publish from the terminal or connect a compatible agent.
  • Access controls. Make work private, visible to the workspace, or public. Public links can also use a password.

The current roadmap lives at derive.to/roadmap and keeps its history at the same URL.

The fastest path. No install.

  1. Go to derive.to
  2. Create an account
  3. Publish your first artifact
You get a library, in-browser publishing, version history, and comments. One container is the whole product: API, web, sign-in, publishing, comments, and the sandboxed viewer, with SQLite and blobs in one volume. Follow the self-hosting quick start to install a digest-pinned release or build the current checkout. It includes secure first-user bootstrap, readiness checks, and a verified first backup. See the deployment guide for Postgres, S3/R2, and cloud hosts.

Want a complete artifact rather than a blank starter? Open or publish one of the official examples : a designed launch page, a research brief, or a living status report. Each shows a different starting point.

npm i -g @derive-to/cli
derive init my-doc --template slides   # templates: md · html · slides
cd my-doc
derive publish                         # versioned URL; the id is saved to derive.json

Connect your coding agent (MCP)

Add to Cursor or one line for Claude Code: claude mcp add --transport http derive https://derive.to/mcp

# Derive is a remote MCP server (OAuth). Connect either client:
claude mcp add --transport http --scope project derive https://derive.to/mcp
codex mcp add derive --url https://derive.to/mcp

# or run a local stdio server (set DERIVE_SERVER; DERIVE_TOKEN for a static bearer):
npx -y @derive-to/mcp

The agent acts at the role you grant. Publish access publishes directly, as a kept and restorable version with the workspace notified; a lower scope suggests changes in comments for a person to apply. Full setup and tool guidance is in packages/mcp/SKILL.md.

One Node container is the whole product; storage is pluggable behind interfaces. The same image self-hosts on SQLite and local disk, scales on Postgres and S3/R2, or runs on Cloudflare Workers.

apps/api          HTTP API, sandboxed artifact serving, viewer
apps/web          web UI (TanStack Start, SPA mode, static bundle)
packages/core     domain: ports, publish, markdown render, viewer shell
packages/db       MetaStore: sqlite (default) · postgres · d1
packages/storage  BlobStore: fs (default) · s3/r2
packages/cli      derive init (md/html/slides) · derive publish <file|dir> · derive runner serve (run an Agent)
packages/mcp      Local compatibility MCP: eight agent tools + derive://guide

Every artifact ships OG and Twitter meta plus an oEmbed document, serves a live Server-Sent Events stream, and renders under a strict sandbox CSP on an opaque origin. See the artifact authoring standard for authoring and embed details.

Functional Source License (FSL-1.1-ALv2), fair source. Run, modify, and self-host Derive freely for any purpose except offering it as a competing commercial product or service. Each release automatically converts to Apache-2.0 two years after it ships. See the plain-English licensing guide.