video-script-to-post

短视频全链路编导引擎，从热点选题→分镜脚本→AI生成提示词→多平台排期，一站式输出可执行内容包

Security Scan

The skill's declared purpose (short-video script + AI prompt + platform strategy) matches its instructions and metadata; it is an instruction-only skill with no installs or credential requests and thus is internally coherent.

Purpose & Capability

Name/description, schema, examples, and SKILL.md all focus on generating short-video storyboards, AI visual prompts, platform-specific publishing guidance, and compliance hints. Nothing in the package requests unrelated capabilities (cloud keys, system access, or external binaries).

Instruction Scope

Runtime instructions are limited to content generation tasks (structure planning, shot breakdown, AI prompt generation, platform strategy, and compliance checks). The only slight vagueness: the '自动检测广告法敏感词/版权风险' step doesn't specify authoritative rule-sources — it will rely on the agent/model's knowledge or external data the agent already has access to, but this is consistent with the skill's purpose.

Install Mechanism

No install spec and no code files (instruction-only). No downloads or package installs are requested, which minimizes disk-write/execution risk.

Credentials

The skill declares no required environment variables, credentials, or config paths. There are no unexpected secret requests or cross-service credentials; the requested inputs are just content parameters (topic, platform, duration, tone).

Persistence & Privilege

always is false and the skill does not request persistent presence or system-wide config changes. Autonomous invocation is allowed (the platform default) but the skill itself does not escalate privileges or modify other skills.

Assessment

This skill is instruction-only and internally consistent with its stated purpose, so the main risks are content/legal rather than technical: (1) Review generated AI prompts and visual outputs for copyrighted material or trademarked imagery before publishing; (2) treat the 'compliance' checks as advisory — have a human/legal reviewer confirm anything flagged; (3) because the package has no owner homepage and comes from an opaque owner ID, prefer testing outputs in a sandbox/qa workflow before using in production; (4) if you plan to wire this skill into external tools or models (Midjourney/Runway/etc.), only provide the minimal, dedicated credentials those services require and review their usage policies.

🎬 短视频全链路编导

🎯 核心定位

将创意构思转化为可直接交付拍摄/AI生成的标准化内容工程包，深度适配主流短视频平台算法逻辑与用户停留习惯。

🔄 工作流指令

1. 结构规划：根据 target_platform 与 tone_style 构建内容骨架（黄金3秒钩子→核心价值→行动号召）。
2. 分镜生成：输出秒级分镜脚本表：画面描述/台词字幕/时长分配/音效BGM/AI视觉提示词。
3. AI参数包：生成适配 Midjourney/ComfyUI/Runway/Pika 的精确提示词与参数配置（比例/风格/运动控制）。
4. 平台策略：输出平台专属优化方案（封面标题/推荐标签/发布时间/评论区互动话术）。
5. 合规校验：自动检测广告法敏感词、版权风险与平台社区公约红线。
6. 模板输出：按标准 Markdown 结构生成工程包。

📤 输出模板

# 🎥 短视频内容工程包

## 1. 分镜脚本表
| 秒数 | 画面描述 | 台词/字幕 | 音效/BGM | 视觉提示词(AI) |
|:---|:---|:---|:---|:---|
| 0-3s | ... | ... | ... | `prompt...` |
| ... | ... | ... | ... | ... |

## 2. AI生成参数包
- **Midjourney/ComfyUI**: `--v 6.0 --ar 9:16 --style raw --s 750 ...`
- **Runway/Pika**: `motion: 7, camera: pan_left, seed: 42, negative_prompt: ...`

## 3. 平台运营策略
| 维度 | 抖音 | B站 | 视频号 |
|:---|:---|:---|:---|
| 封面标题 | ... | ... | ... |
| 推荐标签 | `#...` | `#...` | `#...` |
| 互动引导 | ... | ... | ... |

## 4. 合规与版权提醒
- 字体/音乐版权状态：...
- 广告法敏感词检测：...
> 📌 本包需经编导/法务复核后进入制作流。AI提示词需根据实际模型版本微调参数。