CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

anime-forge

讲角色自己故事的动画短片:先把故事验明白(考证真事、一句话、剧本、冷读),再为最危险那一段定设定、做最小样张;样张过了出全片正式风格的分镜图,拼成动态分镜,最后后期; 只有少数几格要「活起来」时才走图生视频。 Use when: 做目的就是讲角色自己故事的动画短剧、角色 IP 短片、醋醋喵这类猫咖新集;把家里的真事改编成动画; 给已有剧本做设定、声音稿、分镜、动态分镜、精绘或后期。 Not for: 推产品、讲功能的片子,哪怕主角是猫、片里有剧情(用 video-forge); 录屏剪辑(用 video-forge);单张图或单页漫画(用 image-generation);PPT(用 ppt-forge)。 Output: 带真事对照的剧本 md、设定图、声音稿、分镜表、分镜图、动态分镜粗片,以及 Works 视频作品和确定版本的成片卡。

Quality

91%

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

81%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 strong, hard-won process skill: the 11-step workflow has explicit review and spending checkpoints, feedback loops, and a concrete mistakes table, all written for a competent reader without padding. Main weaknesses are a dangling `refs/i2v-route.md` path, heavy dependence on files outside the skill bundle, and a couple of high-level stages (后期) without executable detail.

Suggestions

Fix the dangling reference: `refs/i2v-route.md` is cited twice but no refs/ directory exists in the bundle — move the file into the bundle (e.g. references/i2v-route.md) or correct the path so the i2v route is actually reachable.

Add executable detail to the thinnest stages: step 10 (后期) lists effects (视差、眨眼、盖章震动) without commands or parameters, and step 7's `tts-check.py` is named but never shown with an invocation example.

Tighten the narrative retellings in '先懂一件事' and the 设定先行 '为什么' paragraph — compress the EP0/You anecdotes to their one-line lesson so the earned wisdom stays while the story padding goes.

DimensionReasoningScore

Conciseness

The body is dense with project-specific knowledge Claude cannot infer ('EP0 草图 241 张、精绘 112 张', '一张草图和一张彩图的生成耗时接近') and assumes competence, with the current-vs-old route flagged up front. A few narrative retellings ('前 55 秒每张都好看,连起来看不懂', story quotes) could be trimmed without losing guidance. Not 5: minor padding remains; not 3: nothing explains general concepts.

4 / 5

Actionability

Mostly executable guidance: named scripts to run ('渲染前跑 `../video-forge/scripts/timeline-check.py`'), concrete parameters ('2–4 格画面配好音,10–20 秒'), API usage details ('read-asset 带 expectedRevision', 'assets[].bindingSelection'), and a Common Mistakes table with specific fixes. Not 5: a few stages stay high-level (step 10 后期 lists effects without how-to) and no inline command examples are shown.

4 / 5

Workflow Clarity

The 11-step 主流程 table is clearly sequenced with explicit validation checkpoints (the 'You 看什么' review gates at steps 1, 5, 6, 9, 10), spending gates ('样张过之前,只为这一段花钱'), per-batch visual checks ('每批先把成品的脸和设定并排对一遍再往下'), and error-recovery feedback loops in the Common Mistakes and 交片与改片 sections. Fully matches the top anchor.

5 / 5

Progressive Disclosure

Good structure: the body is an overview with clear sections, the one bundle file (assets/art-brief-header.md) is real and appropriately delegated ('画稿开头抄 `assets/art-brief-header.md`'), and references are one level deep and contextually signaled. Not 5: `refs/i2v-route.md` is referenced but no refs/ directory exists in the bundle, and most supporting detail (docs/, ../.cat-cafe-shared-refs/, ../video-forge/) lives outside the skill, leaving the in-bundle reference structure thin.

4 / 5

Total

17

/

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 multi-stage pipeline concretely, gives explicit 'Use when' triggers with natural synonyms, lists deliverables, and cleanly disambiguates against adjacent skills with a 'Not for' section. Both what and when are unambiguous.

DimensionReasoningScore

Specificity

The description enumerates the full pipeline concretely — '考证真事、一句话、剧本、冷读', '定设定、做最小样张', '分镜图,拼成动态分镜,最后后期', '图生视频' — plus an explicit 'Output:' list of deliverables (剧本 md、设定图、声音稿、分镜表、分镜图、动态分镜粗片、成片卡). Comprehensive coverage of concrete actions with no gaps.

5 / 5

Completeness

Both questions are answered explicitly: 'what' via the pipeline and 'Output:' list, 'when' via a literal 'Use when:' clause with concrete triggers ('做目的就是讲角色自己故事的动画短剧…给已有剧本做设定、声音稿、分镜、动态分镜、精绘或后期'), plus a 'Not for:' exclusion list.

5 / 5

Trigger Term Quality

Natural user phrasing is well covered with synonyms: '动画短剧、角色 IP 短片、醋醋喵这类猫咖新集', '把家里的真事改编成动画', '设定、声音稿、分镜、动态分镜、精绘或后期'. These are exactly the words a user needing this skill would say.

5 / 5

Distinctiveness Conflict Risk

A clear niche (films whose purpose is to tell the character's own story) is paired with explicit routing away from siblings: '推产品、讲功能的片子…(用 video-forge);录屏剪辑(用 video-forge);单张图或单页漫画(用 image-generation);PPT(用 ppt-forge)'. Minimal conflict risk.

5 / 5

Total

20

/

20

Passed

Validation

87%

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

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 2 suspicious

Warning

referenced_paths_exist

Referenced path issues: 1 missing

Warning

Total

14

/

16

Passed

Repository

zts212653/clowder-ai

Path

cat-cafe-skills/anime-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.