CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

sprite-forge

角色动态 sprite / 桌宠皮肤 / atlas 状态行生产线:母图锚身份 → Frame A 首帧 → AI 视频生成(人工环节)→ 截帧确定性加工 → atlas row + 三道闸。 Use when: 做桌宠皮肤、给猫补新状态动画行、visible-cafe/F258 猫形象素材、 角色小循环动画(呼吸/走路/睡觉/挥手级)、把静态角色"做成动态的"。 Not for: 动画短片成片(用 anime-forge)、录屏/教程视频(用 video-forge)、 单张静态图/头像(用 image-generation)、UI 图标。 Output: 192×208 atlas row strip + QA 包(contact/GIF/provenance)→ 过三道闸 + operator identity veto 后接 pet.json。

Quality

96%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

92%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A tightly engineered operational document: complete pipeline with owners and fallbacks, copy-paste deterministic tooling, explicit batch-gate validation with operator veto, and a mistake table where every row cites a real incident. The only slack is a small amount of origin lore and incident dates that could be compressed.

DimensionReasoningScore

Conciseness

The body is dense and assumes competence — no space is spent explaining ffmpeg, video models, or sprite concepts, and heavy material is delegated to external docs ("重材料按需读,不复制进上下文"). Minor trimmable narrative remains: the origin blockquote ("Skill 化诱因:F258 素材首跑两连翻车…") and repeated incident-date citations (2026-06-22~24, 2026-07-08, 2026-07-09) are lore rather than instruction, which is what keeps this below a 5.

4 / 5

Actionability

Step 4 is fully executable with copy-paste commands and expected results in comments ("ffmpeg -i video.mp4 -vf \"fps=4\" frame_%02d.png # 2-3s → 8-12 帧" through the mandatory 59×64 preview), exact cell dimensions are given, and verbatim prompt patterns are quoted ("Only differences: ... Everything else IDENTICAL", the 2D style-anchor first line with its negative list). Fallbacks are concrete: B-frame drift → degrade to single-image i2v; dirty edges → rerun despill.

5 / 5

Workflow Clarity

The five-step table sequences the pipeline with owner, constraints, and fallback per step, and this batch workflow has explicit validation and feedback loops: "每批过了才做下一批(2-2-3 批次制先例)", three gates plus operator identity veto as final review, and recovery guidance such as "抽卡 2-3 败即换姿势写法别硬 roll" and "脏边则重跑 despill". Red line 6 closes the loop by requiring validated prompts to flow back to provenance.

5 / 5

Progressive Disclosure

The body stays a lean overview and pushes deep material one level deep into clearly signaled files under a dedicated section: "完整提示词模板库 + Spike 数据 + 9 态任务表:docs/features/assets/F229/desktop-pet-sprite/video-to-spritesheet-pipeline.md", "近例 prompt pack…prompt-pack-v1.md", the cell spec contract, and a pet.json example. Each reference states what it contains and when to read it; nothing is nested further, and no bundle files exist to inline.

5 / 5

Total

19

/

20

Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

An exemplary description: it states the full production pipeline with a concrete output format, gives explicit Use-when triggers including a colloquial one, and cleanly fences off sibling skills with a Not-for list. Every clause is operational rather than promotional.

DimensionReasoningScore

Specificity

The description lays out the entire pipeline as concrete actions ("母图锚身份 → Frame A 首帧 → AI 视频生成(人工环节)→ 截帧确定性加工 → atlas row + 三道闸") and specifies the exact deliverable ("192×208 atlas row strip + QA 包(contact/GIF/provenance)→ … 后接 pet.json"). Coverage is comprehensive from input anchor to final integration, with the human-only step honestly flagged.

5 / 5

Completeness

Both questions are explicitly answered: 'what' via the pipeline and Output clauses, 'when' via the explicit "Use when:" list with concrete trigger phrases. It goes further with a "Not for:" clause routing adjacent requests to sibling skills.

5 / 5

Trigger Term Quality

The Use-when clause lists natural phrasings a user would actually say: "做桌宠皮肤", "给猫补新状态动画行", "角色小循环动画(呼吸/走路/睡觉/挥手级)", and the colloquial "把静态角色\"做成动态的\"". Domain vocabulary (sprite, 桌宠皮肤, atlas 状态行) plus these colloquial variants gives comprehensive keyword coverage.

5 / 5

Distinctiveness Conflict Risk

The niche is unambiguous (looping state-row sprite assets for desktop pets) and the Not-for clause explicitly disambiguates against the nearest neighbors: "动画短片成片(用 anime-forge)、录屏/教程视频(用 video-forge)、单张静态图/头像(用 image-generation)、UI 图标". Minimal conflict risk.

5 / 5

Total

20

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

zts212653/clowder-ai

Path

cat-cafe-skills/sprite-forge/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.