跳到正文
cccskills
免费在 GitHub 公开

img2mo-std

Standardize a 2D animation baseline image with `/img2mo-std xxx.png/pos`. Use when preparing a character, creature, prop, or weapon baseline before motion generation, especially if the source image is too large, tightly cropped, lacks transparent action margin, has a white background, or later walk/attack/idle frames show clipping, overlap, scale popping, or inconsistent character size.

中文简介正在准备中。当前显示原文说明。

查看安装方法

包含的文件(1)

  • SKILL.md3.3 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Img2mo-std

Overview

Use this skill when the user invokes:

/img2mo-std xxx.png/pos

The command audits and standardizes a baseline frame before animation generation. It creates a smaller, padded, transparent-canvas source image that leaves enough room for walk, attack, idle, weapon swing, cape, tail, horn, or limb stretch poses.

Input Resolution

Resolve the argument after /img2mo-std as follows:

  1. If it is an existing relative or absolute path, use it directly.
  2. If it is a bare file name such as s7.png, first try sample\s7.png.
  3. If it has no extension such as s7, first try sample\s7.png, then sample\s7.jpg, then sample\s7.webp.
  4. If no matching image exists, report the missing path and do not guess from unrelated files.

Examples:

/img2mo-std s7
/img2mo-std s7.png
/img2mo-std sample\s7.png
/img2mo-std C:\AI\2dimg2motion-v1\sample\s7.png

Audit Criteria

A baseline is not standard if any of these are true:

  • The foreground subject is larger than about 300-400 px on its longest side.
  • The subject touches or nearly touches the canvas edge.
  • There is not enough empty transparent space for arms, legs, weapons, tails, capes, horns, or effects to extend during motion.
  • The background is opaque white or near-white when the character should be isolated.
  • The source has a huge canvas that will slow generation without adding useful detail.
  • Previous generated frames show inconsistent character size, off-center motion, clipped body parts, or crowded sprite-sheet cells.

Tool

Run the repository tool from the workspace root after resolving the image path:

python scripts\standardize_baseline.py <resolved-image-path>

Default output:

sample\<input-stem>-standard.png

Useful options:

python scripts\standardize_baseline.py sample\s7.png --subject-max 360 --margin-ratio 0.75
python scripts\standardize_baseline.py sample\s7.png --output sample\s7-standard.png
python scripts\standardize_baseline.py sample\s7.png --check-only

Recommended default target:

  • Subject longest side: 360
  • Margin ratio: 0.65 to 0.85
  • Output format: PNG with transparency

Report Interpretation

The tool prints a JSON report. Treat the standardized output as ready when:

  • needs_standardization is false after processing.
  • subject_longest_side is between 300 and 400.
  • edge_clearance_ratio leaves enough room for the intended motion.
  • warnings is empty or only contains acceptable notes.

If the report still warns that the image is tight or too large, rerun with a smaller --subject-max or larger --margin-ratio.

Animation Workflow Rule

After /img2mo-std, use the standardized output image as the baseline for all later walk, attack, idle, or other sprite generation. Do not continue from the original oversized or tightly cropped source unless the user explicitly asks to bypass standardization.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Use when a user provides one baseline image and requests game animation frames, sprite sequences, attack/walk/idle/hit/death/casting motion, transparent PNG frames, a spritesheet, keyframe prompts, or consistent whole-character pose animation.

中文简介正在准备中。当前显示原文说明。

WU-HAOTIAN34/2dimg2motion2272026年7月2日 更新

Learn reusable 2D motion-generation knowledge from user-specified action resources with `/img2mo-learn <resource>`. Use when the user provides videos, extracted frame sequences, spritesheets, Spine assets, generated outputs, failed attempts, or reference motion folders and wants to summarize animation timing, pose beats, style traits, prompt patterns, extraction/cropping rules, or failure lessons into the project `img2mo-knowledge/` folder for later `/img2motion` generation.

中文简介正在准备中。当前显示原文说明。

WU-HAOTIAN34/2dimg2motion2272026年7月2日 更新

查看 WU-HAOTIAN34 的全部技能

报告此技能的问题