img2mo-learn

Name: img2mo-learn
Author: WU-HAOTIAN34

101stars2forks

How to install img2mo-learn?

wu-haotian34/2dimg2motion/img2mo-learn

$npx -y skills add WU-HAOTIAN34/2dimg2motion --skill img2mo-learn

Installs into the current project.

Summary

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

Files · 1

View on GitHub

SKILL.md

1	# Img2mo-learn
2
3	## Overview
4
5	Use this skill when the user invokes:
6
7	```text
8	/img2mo-learn <resource-path-or-folder>
9	img2mo-learn <resource-path-or-folder>
10	```
11
12	The goal is to turn finished or reference motion assets into reusable project knowledge. Store learned knowledge in the project-level `img2mo-knowledge/` folder, never in the installed Codex skill directory during normal work.
13
14	## Input Resolution
15
16	Resolve the argument after `img2mo-learn` as follows:
17
18	1. If it is an existing relative or absolute path, use it directly.
19	2. If it is a bare name, first try `sample\<name>`, then `output\<name>`, then `motion\<name>`.
20	3. If it is a folder, inspect likely assets in this order: `manifest.json`, `preview.gif`, `contact-sheet.`, `spritesheet.`, `fullframe/`, `frames/`, Spine `.json/.atlas/.skel`, then videos.
21	4. If no matching resource exists, report the missing path and ask for the correct path.
22
23	Supported resources:
24
25	- video files such as `.mp4`, `.mov`, `.webm`;
26	- PNG frame folders, spritesheets, contact sheets, or GIF previews;
27	- project outputs from this skill such as `output/<action-id>/`;
28	- Spine-style assets such as `.json`, `.atlas`, `.skel`, texture folders;
29	- local reference-library folders under `motion/`.
30
31	## Knowledge Location
32
33	Create this structure if missing:
34
35	```text
36	img2mo-knowledge/
37	\|-- index.md
38	\|-- learnings.jsonl
39	\|-- action-patterns.md
40	\|-- style-patterns.md
41	\|-- prompt-patterns.md
42	`-- failures.md
43	```
44
45	Append one JSON object per learning session to `img2mo-knowledge/learnings.jsonl`. Keep Markdown files concise and curated; do not paste huge logs, full prompts, or complete frame listings.
46
47	## Learning Workflow
48
49	1. Identify the resource type.
50	- For video: read frame size, fps, duration, and frame count with `ffprobe` when available.
51	- For frame sequences: count frames, inspect canvas sizes, alpha/background, and contact sheet if present.
52	- For spritesheets: infer grid/cell count when possible; otherwise describe visible beats.
53	- For Spine assets: inspect animation names, bone/slot names, skins, attachments, timeline names, and texture organization without assuming rendered motion if frames are not available.
54
55	2. Create review surfaces if useful.
56	- For video or frame folders, create temporary or output-side contact sheets and preview GIFs if they do not exist.
57	- Do not alter the source resource.
58	- Do not store bulky extracted frames in `img2mo-knowledge/`; store outputs under `output/` or `tmp/` and reference their paths in JSON.
59
60	3. Summarize motion timing.
61	- Identify action type: attack, walk, idle, block, suffer, death, born, skill/cast, or other.
62	- Record frame count, fps, loop behavior, and major beats.
63	- For attacks, prefer beat labels such as `guard`, `anticipation`, `acceleration`, `contact`, `contact hold`, `follow-through`, `recovery`.
64	- Record which frame ranges are most useful as key poses.
65
66	4. Summarize pose and topology lessons.
67	- Record active limb/feature, weapon or prop owner, anchor limb/surface, facing direction, and stable baseline behavior.
68	- Note silhouette expansion, squash/stretch, center drift, foot/bottom baseline, and whether motion needs extra canvas margin.
69
70	5. Summarize style lessons.
71	- Record line weight, palette, shading, material treatment, outline softness, shape language, effects style, and background/keying considerations.
72	- Distinguish character style from detached effects.
73
74	6. Summarize prompt lessons.
75	- Write reusable prompt clauses that could improve later generation.
76	- Keep prompt clauses short and parameterized; avoid overfitting to one character name unless the lesson is character-specific.
77
78	7. Summarize failures and constraints.
79	- Record what should be rejected: hand swaps, scale popping, bad alpha, cut weapons, over-crowded sheets, text/watermark contamination, or mismatched style.
80
81	8. Write project knowledge.
82	- Append structured session data to `learnings.jsonl`.
83	- Update the relevant Markdown files with durable, reusable lessons.
84	- If the learning is only useful for one output, also write `output/<action-id>/retro.md`.
85
86	## JSONL Schema
87
88	Each line in `img2mo-knowledge/learnings.jsonl` should be a compact JSON object:
89
90	```json
91	{
92	"id": "learn-YYYYMMDD-HHMMSS-short-name",
93	"date": "YYYY-MM-DD",
94	"source": "relative/or/absolute/path",
95	"resource_type": "video\|frame_sequence\|spritesheet\|spine\|output\|reference_folder",
96	"action_type": "attac

People who liked this also viewed

6 picks

Type

img2mo-learn

101stars2forks

View on GitHub

How to install img2mo-learn?

wu-haotian34/2dimg2motion/img2mo-learn

$npx -y skills add WU-HAOTIAN34/2dimg2motion --skill img2mo-learn

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Img2mo-learn
2
3	## Overview
4
5	Use this skill when the user invokes:
6
7	```text
8	/img2mo-learn <resource-path-or-folder>
9	img2mo-learn <resource-path-or-folder>
10	```
11
12	The goal is to turn finished or reference motion assets into reusable project knowledge. Store learned knowledge in the project-level `img2mo-knowledge/` folder, never in the installed Codex skill directory during normal work.
13
14	## Input Resolution
15
16	Resolve the argument after `img2mo-learn` as follows:
17
18	1. If it is an existing relative or absolute path, use it directly.
19	2. If it is a bare name, first try `sample\<name>`, then `output\<name>`, then `motion\<name>`.
20	3. If it is a folder, inspect likely assets in this order: `manifest.json`, `preview.gif`, `contact-sheet.`, `spritesheet.`, `fullframe/`, `frames/`, Spine `.json/.atlas/.skel`, then videos.
21	4. If no matching resource exists, report the missing path and ask for the correct path.
22
23	Supported resources:
24
25	- video files such as `.mp4`, `.mov`, `.webm`;
26	- PNG frame folders, spritesheets, contact sheets, or GIF previews;
27	- project outputs from this skill such as `output/<action-id>/`;
28	- Spine-style assets such as `.json`, `.atlas`, `.skel`, texture folders;
29	- local reference-library folders under `motion/`.
30
31	## Knowledge Location
32
33	Create this structure if missing:
34
35	```text
36	img2mo-knowledge/
37	\|-- index.md
38	\|-- learnings.jsonl
39	\|-- action-patterns.md
40	\|-- style-patterns.md
41	\|-- prompt-patterns.md
42	`-- failures.md
43	```
44
45	Append one JSON object per learning session to `img2mo-knowledge/learnings.jsonl`. Keep Markdown files concise and curated; do not paste huge logs, full prompts, or complete frame listings.
46
47	## Learning Workflow
48
49	1. Identify the resource type.
50	- For video: read frame size, fps, duration, and frame count with `ffprobe` when available.
51	- For frame sequences: count frames, inspect canvas sizes, alpha/background, and contact sheet if present.
52	- For spritesheets: infer grid/cell count when possible; otherwise describe visible beats.
53	- For Spine assets: inspect animation names, bone/slot names, skins, attachments, timeline names, and texture organization without assuming rendered motion if frames are not available.
54
55	2. Create review surfaces if useful.
56	- For video or frame folders, create temporary or output-side contact sheets and preview GIFs if they do not exist.
57	- Do not alter the source resource.
58	- Do not store bulky extracted frames in `img2mo-knowledge/`; store outputs under `output/` or `tmp/` and reference their paths in JSON.
59
60	3. Summarize motion timing.
61	- Identify action type: attack, walk, idle, block, suffer, death, born, skill/cast, or other.
62	- Record frame count, fps, loop behavior, and major beats.
63	- For attacks, prefer beat labels such as `guard`, `anticipation`, `acceleration`, `contact`, `contact hold`, `follow-through`, `recovery`.
64	- Record which frame ranges are most useful as key poses.
65
66	4. Summarize pose and topology lessons.
67	- Record active limb/feature, weapon or prop owner, anchor limb/surface, facing direction, and stable baseline behavior.
68	- Note silhouette expansion, squash/stretch, center drift, foot/bottom baseline, and whether motion needs extra canvas margin.
69
70	5. Summarize style lessons.
71	- Record line weight, palette, shading, material treatment, outline softness, shape language, effects style, and background/keying considerations.
72	- Distinguish character style from detached effects.
73
74	6. Summarize prompt lessons.
75	- Write reusable prompt clauses that could improve later generation.
76	- Keep prompt clauses short and parameterized; avoid overfitting to one character name unless the lesson is character-specific.
77
78	7. Summarize failures and constraints.
79	- Record what should be rejected: hand swaps, scale popping, bad alpha, cut weapons, over-crowded sheets, text/watermark contamination, or mismatched style.
80
81	8. Write project knowledge.
82	- Append structured session data to `learnings.jsonl`.
83	- Update the relevant Markdown files with durable, reusable lessons.
84	- If the learning is only useful for one output, also write `output/<action-id>/retro.md`.
85
86	## JSONL Schema
87
88	Each line in `img2mo-knowledge/learnings.jsonl` should be a compact JSON object:
89
90	```json
91	{
92	"id": "learn-YYYYMMDD-HHMMSS-short-name",
93	"date": "YYYY-MM-DD",
94	"source": "relative/or/absolute/path",
95	"resource_type": "video\|frame_sequence\|spritesheet\|spine\|output\|reference_folder",
96	"action_type": "attac

People who liked this also viewed

6 picks

Type