Social Video Resizer

Adapt one social video into platform-ready aspect ratios and dimensions for TikTok, Reels, Shorts, feeds, stories, and ads without awkward stretching or lost...

Security Scan

The skill is an instruction-only helper for formatting videos to platform aspect ratios; it asks for no credentials, installs, or unrelated access and is internally consistent with its stated purpose.

Purpose & Capability

Name, description, and required inputs (source video, target platforms, priority areas) align with a video-resizing/adaptation helper. It does not request unrelated credentials, binaries, or config paths.

Instruction Scope

SKILL.md contains focused, high-level workflow steps (decide crop/pad/scale, protect faces/subtitles, output versions). It does not instruct reading unrelated files or env vars. Note: it is procedural guidance only and does not specify tools or where processing happens (local vs remote), so implementers must decide how to run the transforms.

Install Mechanism

No install spec and no code files — instruction-only — so nothing is written to disk or downloaded during install.

Credentials

No environment variables, credentials, or config paths are requested. The declared inputs are proportional to the task (a source video and resize targets).

Persistence & Privilege

always is false and the skill is user-invocable; it does not request permanent presence or change other skills' configs. Autonomous invocation is allowed by default but not combined with other red flags.

Assessment

This skill is an instruction-only recipe for producing platform-ready video variants and is internally coherent, but consider the following before use: (1) Clarify where processing occurs — local machine, hosted service, or third-party API — and avoid uploading sensitive footage to unknown endpoints. (2) Confirm which tools will actually perform the resize (e.g., ffmpeg or an editing service) and whether those tools require installation or credentials. (3) Limit the agent's file access to only the source video(s) needed and verify outputs are stored where you expect. (4) If you need automated processing, ask the skill author or integrator to provide an explicit, auditable implementation that documents any network endpoints, binaries used, and data retention policy. These checks will reduce privacy and data-exfiltration risk.

把一条社媒视频，快速变成多个平台可发的尺寸版本。

Problem it solves

同一条内容发不同平台时，尺寸要求经常不一样：

• TikTok / Reels / Shorts 偏 9:16；
• Feed、广告位、封面预览可能要 1:1、4:5、16:9；
• 直接硬拉伸会让人物变形，直接裁切又容易切掉字幕、脸或产品。

这个 skill 的目标是： 在不破坏主体和关键信息的前提下，把一条视频适配成多个社媒平台可直接使用的尺寸版本。

Use when

• 一条视频要同时发 TikTok、Instagram、YouTube Shorts、Facebook、X 等平台
• 需要在 9:16、1:1、4:5、16:9 等比例之间切换
• 需要决定 pad、crop 还是 scale 的最佳策略

Do not use when

• 需要高级主体追踪或复杂智能重构镜头
• 视频内容本身构图太差，需要重新剪辑而非单纯改尺寸

Inputs

• Source video file
• Target platforms or ratios
• Whether the priority is face, product, subtitles, or full-frame preservation
• Optional output dimensions, file size limit, or ad placement requirements
• Optional brand constraints such as background color or padding preference

Workflow

1. 确认目标平台和尺寸要求。
2. 判断该用 crop、pad 还是 scale。
3. 优先保护人物、产品和字幕安全区域。
4. 输出一个或多个平台适配版本。
5. 说明每个版本的构图取舍和兼容性。

Output

Return:

1. Target platform versions
2. Resize strategy per version
3. Output dimensions
4. Framing / subtitle safety notes
5. Additional recommended variants if useful

Quality bar

• 绝不拉伸变形人物或产品
• 重要字幕和产品展示尽量不被切掉
• 优先使用平台安全尺寸
• 清楚说明 crop 与 pad 的取舍

Resource

See references/output-template.md.