dbs-content-risk-check
Contentdontbesilent2025/dbskillskills.sh ↗
Installs
3,586
deduplicated, at the last sync
Since we started
+15.8%
19 readings, about 2 hours apart. Not a live curve.
Our category
Content
ours
Last read
Sep 3, 2026
from the directory
Our brief
oursThis skill functions as a content publishing risk checker for creators, analyzing various elements such as titles, body text, videos, images, and account profiles to identify potential issues that might trigger platform machine moderation or reveal substantive policy violations [1]. It provides specific locations within the content and suggests minimal modifications to mitigate risks while preserving the creator's original style and viewpoint [1].
- Natural language risk report
- Minimal modification actions
- Three lines of video overlay text
- Draft text
- Images
- Videos
- Comments
- Account profiles
not detected
The evidence does not contain information regarding paid plans or API costs.
not detected
The evidence does not mention any registration requirements.
The skill cannot make the final decision on whether content can be published, nor does it default to rewriting entire sections of text [1]. Furthermore, if provided materials are incomplete (e.g., only text), it will remind the user that other areas like images or comments have not been checked [1].
Evidenceskills/dbs-content-risk-check/SKILL.md:1-337
--- name: dbs-content-risk-check description: 面向创作者的内容发布风险检查。逐句检查自媒体、社交媒体和内容平台的标题、正文、图片文字、字幕、口播、评论、账号资料及视频画面,分别判断平台机器可能识别什么、内容本身存在什么实质问题,并给出具体位置和最小修改动作;需要明确视频语境边界时,额外生成真实准确的三行贴片小字。用户询问「检查敏感词」「发布前排雷」「这篇内容哪里可能违规」「这条小红书能不能发」「视频上放什么声明」时使用。保留强观点、情绪张力、个人风格和传播钩子。 --- # dbs-content-risk-check:内容发布风险检查 帮助创作者在发布前找到可能导致拦截、复核、限流或处罚的具体内容,并完成必要的最小修改。默认输出普通创作者可以直接使用的自然语言报告,不替用户作出发布决定,也不默认重写全文。 ## 核心原则 同时回答两个问题: 1. **平台机器可能识别什么**:广告结构、联系方式、外链、二维码、连续数字、敏感词面、OCR 噪声和重复模式可能触发自动审核; 2. **内容本身需要处理什么**:根据完整语义判断欺诈、虚假宣传、违规导流、受限推广、医疗功效、侵权、隐私、攻击和危险行为等实质问题。 两层结论分别成立。机器误伤只能写成「可能触发机器审核」,不能补造合理违规原因。实质问题必须说明独立机制,不能只引用机器标签或关键词。 每一条建议都要帮助创作者完成以下动作: - 在原文或画面中找到具体位置; - 看懂可能产生什么后果; - 知道删什么、改什么、补什么或改用什么官方入口; - 知道哪些强观点、口语和叙事可以保留。 ## 工作任务 用户提交文稿、图片、视频、评论、主页或账号资料后,完成发布前检查,输出自然语言结论和最小修改动作。用户没有要求代改全文时,停在风险定位和局部修改建议。 ## 审核流程 ### 第一步:识别材料和发布情境 根据现有材料识别: - 目标平台; - 内容位置:标题、正文、OCR、字幕、口播、评论、主页或直播话术; - 内容目的:个人记录、观点表达、知识科普、招聘、接单、商品推广、品牌合作或交易; - 发布主体、利益关系、商品或服务类别; - 会改变结论的资质、授权、证据、活动期限和报备情况。 材料不完整时先完成能够完成的检查: - 只有文本:结尾简短提醒图片、主页和评论区尚未检查; - 只有 OCR:把文字识别与原始画面分开,不能用 OCR 结果证明画面安全; - 只有图片:检查文字、人物、动作、物品、标识、隐私和二维码; - 视频:至少检查封面、开头、关键转折、产品特写、结尾引导、字幕和完整音轨; - 平台后台信号不可见:仅列为检查边界,不能反推账号、设备或网络信誉。 商业属性、产品类别或资质可能改变结论时,先给条件化判断,最多补问 1 个
Read 1 of 1 text files in the skill.
Installfrom the directory
npx skills add https://github.com/dontbesilent2025/dbskillInstalling happens there, not here. We are an index with an opinion, not a mirror.
What is inside itfrom the directory
2 files — names only. The directory does not report sizes.
What the auditors foundfrom the directory
A skill is instructions your agent will follow and scripts it may run, so who checked it matters as much as how many people installed it.
Installs, reading by readingours
Axis starts at 3.2k, not zero — the range is 3.2k to 3.6k.
Asked out loudspoken, not typed
The same skill in the words people use speaking to an assistant rather than typing into a box. Each one carries the situation it came from, and each answer says only what the skill's own files support.
The provided evidence does not mention any payment requirements or costs associated with running the skill [1].
No, the skill is designed to provide localized modification suggestions and risk reports, but it will not default to rewriting your entire text [1].
If you only provide an image, the skill will check for text, people, actions, objects, identifiers, privacy issues, and QR codes within that picture [1].