No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →
by naive-kun · Codex Skill · ★ 92
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
🔒 Is naive-video-skill safe to install? View the security audit →
Naive Video Skill 给 Codex 使用的开源口播视频成片导师工作流:原片可以先走可选粗剪,也可以从现成粗剪或字幕轴开始,再完成素材插入、风格选择、可选 ShotCraft 镜头参考、HyperFrames + GSAP 预览和最终成片。 它不只是一份提示词。它包含首次初始化、可选 粗剪、阶段路由、项目状态、体检、迁移、质量闸门和显式反馈学习,目标是像导师一样一次问一个问题,让完全没有剪辑基础的人也能一步步拿到可播放成片。 三步开始 下载并安装 安装后重启 Codex,或新建一个任务。 WorkBuddy 用户 新版本整个仓库只暴露一个 ,所以 WorkBuddy 新安装时只应出现一个技能: 初始化、粗剪、字幕、设计、预览和导出都是这个技能内部的工作流,不会再显示成十几个独立开关。如果旧版已经出现多个 技能,请在 WorkBuddy 里停用或卸载这些旧条目,只保留 ;更新后的仓库不会再次创建它们。 在素材目录打开 Codex 把原视频放在任意本地目录,然后在该目录打开 Codex。你不需要先创建工程。 只说这一句 Skill 会检查环境、读取视频参数、询问最少量的风格问题,并告诉你下一句该说什么。你可以什么都不准备直接用默认风格,也可以补一张喜欢的截图,让 Skill 参考它的配色、层级、卡片和构图。 新手会经历什么 默认不改原视频、不改变主音频时钟;只有你明确同意粗剪策略后,才会把新生成的粗剪版作为后续工作时钟。它也不会用低清代理冒充最终成片,或让截图被字幕和卡片盖住。 原片也能开始:可选 Video Use 粗剪 如果你拿来的是多次重拍、带口误和长停顿的原片,Skill 会先问: 已经粗剪好:直接进入字幕和设计,不会拿一堆转写选项打断你。 还没粗剪:可选调用独立的 Skill,先盘点素材、提出保留/删除策略,得到你确认后再生成一个不覆盖原片的粗剪版。 没安装 :会告诉你它是可选能力,并在你同意后再教你安装;不会偷偷装依赖。 需要靠转写找口误和剪切点时,会说明两条路,但不强制选择: 云端词级转写:精准粗剪的推荐路线,尤其适合口误、重复和多 take;
| Stars | 92 |
| Forks | 12 |
| Language | Python |
| Category | Codex Skill |
| License | MIT |
| Quality Score | 62.2473833351415/100 |
| Last Updated | 2026-08-11 |
| Created | 2026-07-06 |
| Platforms | codex, python |
| Est. Tokens | ~15k |
Explore other popular codex skill tools:
naive-video-skill is A Codex skill for turning talking-head videos into captioned, animated final videos.. It is categorized as a Codex Skill with 92 GitHub stars.
naive-video-skill is primarily written in Python.
You can find installation instructions and usage details in the naive-video-skill GitHub repository at github.com/naive-kun/naive-video-skill. The project has 92 stars and 12 forks, indicating an active community.
naive-video-skill is released under the MIT license, making it free to use and modify according to the license terms.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: