user-content-safety

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

user-content-safety

The moment strangers can put content in front of other strangers, the site needs a way to hear about abuse and a way to act on it, and its rendering must not let content attack visitors.

一旦陌生人可以把内容展示给其他陌生人，网站就需要有渠道接收滥用举报并采取处置措施，同时其渲染机制不得让内容攻击访问者。

Activate when

启用条件

Posts, comments, reviews, profiles with bios/avatars, uploads, messages, public galleries, or shared links exist.
Not when users only see their own private data, or when content is admin-authored only.

存在帖子、评论、评论文章、包含个人简介或头像的个人资料、上传内容、消息、公开图库或共享链接。
如果用户只能看到自己的私人数据，或者内容完全由管理员创作，则不适用。

Working alone

独立工作

This skill is self-contained. Its companions (declared in its metadata) are skills whose method it may need to do its own promised work. Use of a companion can be conditional: declaring one does not mean running it. When a companion's lane applies, use the skill if it is installed; if not, follow its short entry in references/companion-methods.md and say in your report which lanes ran inline at reduced depth. Never skip an applicable lane silently. Skills mentioned here only for escalation, referral, documentation, or optional deeper follow-up are not dependencies: report the hand-off and finish honestly.

Companions:

design-system-reconnaissance

此 skill 是自包含的。它的 companions（在其 metadata 中声明）是指它可能需要调用其方法来完成自身承诺工作的 skills。使用 companion 可以是有条件的：声明一个 companion 并不意味着一定会运行它。当某个 companion 的工作范围适用时，如果该 skill 已安装，请使用它；如果尚未安装，则遵循 references/companion-methods.md 中的简要说明，并在报告中说明哪些工作范围以内联方式按较低深度执行。绝不静默跳过任何适用的工作范围。此处仅针对升级、转介、文档记录或可选深入跟进而提及的 skills 不属于依赖项：应报告交接情况，并如实完成工作。

Companions:

design-system-reconnaissance

。

Inspect

检查

Where content appears and to whom (public, logged-in, group, private link). Anonymous posting? Signup friction?
Reporting: a visible way to report content or a user; the report goes somewhere real (email inbox, queue) and is monitored; confirmation is shown.
Moderation hooks: can the owner remove content and suspend users (
```
admin-authorization
```
,
```
admin-dashboard
```
only if none exists and it is needed)? Are actions logged (
```
admin-audit-log
```
)? Any automated filtering the copy claims?
Abuse controls: rate limits on posting, signup, and uploads; spam prevention (captcha with privacy trade-offs noted); link/URL handling (
```
rel="ugc nofollow"
```
); blocking.
Rendering safety: user text rendered as text or sanitized (XSS:
```
dangerouslySetInnerHTML
```
,
```
innerHTML
```
, markdown renderers without sanitization); user-supplied URLs in
```
href
```
/
```
src
```
(javascript: URLs); uploads validated by type/size, served from a separate origin or with safe headers, not executable; image metadata (EXIF location) stripped where relevant.
Privacy of contributors: public display of email/full name by default; deletion of authored content (
```
data-rights
```
).
Rules and enforcement: community rules or acceptable-use text exist and match what moderation can do (
```
terms-of-service
```
).
Takedown/IP contact for infringement reports (
```
legal-identity-notices
```
).
Children: if minors may post, escalate (
```
minors-readiness
```
).

内容出现在哪里以及面向谁（公开、登录后、群组、私有链接）。允许匿名发布吗？注册门槛如何？
**举报：**有可见的举报内容或用户的方式；举报会进入真实目的地（电子邮件收件箱、队列）并受到监控；会显示确认信息。
**审核钩子：**产品所有者能否删除内容并停用用户（仅当不存在且确有需要时，才使用
```
admin-authorization
```
、
```
admin-dashboard
```
）？操作是否被记录（
```
admin-audit-log
```
）？文案是否声称存在任何自动过滤功能？
**滥用控制：**对发布、注册和上传设置速率限制；垃圾内容防护（captcha，同时说明隐私权衡）；链接或 URL 处理（
```
rel="ugc nofollow"
```
）；屏蔽机制。
**渲染安全：**用户文本按文本渲染或经过清理（XSS：
```
dangerouslySetInnerHTML
```
、
```
innerHTML
```
、未进行清理的 markdown 渲染器）；用户提供并用于
```
href
```
或
```
src
```
的 URL（
```
javascript:
```
URL）；上传文件按类型和大小进行验证，从独立源提供或附加安全 headers，且不可执行；在相关场景中移除图片元数据（EXIF 位置信息）。
**贡献者隐私：**默认公开显示电子邮件或全名；删除所创作的内容（
```
data-rights
```
）。
**规则与执行：**存在与实际审核能力相符的社区规则或可接受使用文本（
```
terms-of-service
```
）。
下架或 IP 联系人，用于受理侵权举报（
```
legal-identity-notices
```
）。
**儿童：**如果未成年人可以发布内容，则升级处理（
```
minors-readiness
```
）。

Evidence that counts

有效证据

Label each claim OBSERVED, SOURCE-INDICATED, DECLARED, INFERRED, UNKNOWN, or REVIEW REQUIRED. UNKNOWN is never a pass and never a failure.

Unsafe rendering is SOURCE-INDICATED until you prove it on staging with an inert test payload (e.g. text containing
```
<b>x</b>
```
renders as text).
A "Report" button that goes nowhere is a finding; test that the report arrives.

将每项主张标记为 OBSERVED、SOURCE-INDICATED、DECLARED、INFERRED、UNKNOWN 或 REVIEW REQUIRED。UNKNOWN 既不表示通过，也不表示失败。

在 staging 上使用惰性测试载荷证明之前，不安全渲染应标为 SOURCE-INDICATED。例如，包含
```
<b>x</b>
```
的文本被渲染为文本。
一个“举报”按钮如果无实际去向，就是一项发现；应测试举报是否确实送达。

May change

可以更改

Design first. Before creating or changing anything visible, inspect the project's existing design system (

design-system-reconnaissance

) and build from its tokens and components, by the component ladder: reuse, compose, extend, and only then create a matching component. Never impose a ReadyVibe look on the user's site.

Add/repair a report link to an existing inbox or form; sanitize or escape rendering; add

rel="ugc nofollow noopener"

; add basic rate limits where the project has a pattern; tighten upload validation. Do not build a moderation platform, invent an inbox, or claim moderation exists.

**设计优先。**在创建或更改任何可见内容之前，先检查项目现有的设计系统（

design-system-reconnaissance

），并依据其 design tokens 和 components 进行构建，按照 component ladder 的顺序执行：reuse、compose、extend，只有在前述方式均不适用时，才创建匹配的 component。绝不能将 ReadyVibe 风格强加到用户的站点上。

将举报链接添加至现有收件箱或表单，或修复已有链接；对渲染进行清理或转义；添加

rel="ugc nofollow noopener"

；在项目已有相应模式的地方添加基础速率限制；收紧上传验证。不要构建审核平台，不要虚构收件箱，也不要声称存在审核机制。

Must not claim

不得声称

"Safe", "moderated", "abuse-free", "we review all content", or anything about legal immunity or liability.

“安全”、“已受审核”、“无滥用”、“我们会审核所有内容”，或任何有关法律豁免或责任的说法。

Verify

验证

On staging: post inert HTML/script-like text and confirm it is inert; submit a report and confirm delivery; hit the posting endpoint repeatedly to confirm limits; upload a disallowed type and confirm rejection.

在 staging 上：发布惰性的类 HTML 或脚本文本，并确认其保持惰性；提交举报并确认送达；反复调用发布 endpoint 以确认限制；上传不允许的类型并确认被拒绝。

Escalate

升级处理

Public posting by minors, adult content, harassment risk, legal-content categories (hate, exploitation), and any real report of harm: human review and, for serious cases, legal counsel.

未成年人公开发布内容、成人内容、骚扰风险、涉及法律问题的内容类别（仇恨、剥削），以及任何真实发生的伤害报告：应交由人工审核；严重案件还应交由法律顾问处理。

No change is valid when

无需更改的有效情形

The product has no user-visible user content. Record the routes and forms checked.

产品没有用户可见的用户内容时，不做更改也是有效结果。记录已检查的 routes 和 forms。