muapi-ai-fight-scene

Generate a high-cut-density action / fight scene by first composing a 16-cell storyboard image, then driving Seedance 2.0 image-to-video off that storyboard. Stacks GPT-Image-2 (character sheet + storyboard), Nano-Banana-2 (environment concept), and Seedance 2.0 i2v.

Quality

60%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./library/motion/ai-fight-scene/SKILL.md

Quality

Content

71%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.

The content is a strong, actionable recipe with concrete prompts, named models, clear phased sequencing, and built-in regeneration feedback loops; it stays lean and avoids over-explaining known concepts. The main gap is progressive disclosure: everything is inlined in one file with no supporting reference files split out, which is acceptable for a self-contained skill but leaves structure at the midpoint.

Suggestions

If the skill grows, move the full per-phase prompt templates or the 'Notes for the Executing Agent' CLI/curl details into a single referenced file (e.g. references/prompts.md) and signal it from the body.

Add one explicit validation gate before the expensive Phase D video run (e.g. 'Only proceed to video generation once the storyboard is approved'), to lift workflow clarity to the top anchor.

Confirm the multi-reference field name for `gpt-image-2-image-to-image` (currently described as 'or the model's documented multi-ref field') so the Phase C step is fully copy-paste ready.

Dimension	Reasoning	Score
Conciseness	The body is largely efficient — concrete prompts, model IDs, and an ASCII pipeline diagram earn their place without restating concepts Claude already knows — with only minor sections (e.g. the 'Notes' rationale paragraphs and the duplicated bold restatement of the core idea under the H1) that could be trimmed.	4 / 5
Actionability	It gives concrete, near-executable guidance per phase: named models (`gpt-image-2-text-to-image`, `nano-banana-2`, `seedance-v2.0-i2v`), full prompt templates with `{{input}}` placeholders, aspect ratios, and a curl/poll fallback, with only small gaps like the exact `images_list` field name flagged as 'documented'.	4 / 5
Workflow Clarity	The four phases (A–D) are clearly sequenced with explicit approval/confirmation checkpoints after each phase and feedback loops ('If a panel reads poorly, regenerate...'; regenerate Phase D first), missing only a fully explicit validate-before-proceed gate to reach the top anchor.	4 / 5
Progressive Disclosure	There are no bundle files (references/scripts/assets) and the SKILL.md is a single well-sectioned document, but all detail — full prompt templates, pipeline notes, and agent-execution notes — is inlined rather than split into one-level-deep referenced files, so structure is present but content that could be separate is inline.	3 / 5
	Total	15 / 20 Passed

Description

50%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.

The description is specific and names concrete actions plus the model stack, but it is missing any explicit 'when to use' trigger guidance, which caps its completeness and leaves trigger coverage only moderate. Adding a 'Use when...' clause with natural terms like 'action sequence', 'fight scene', and 'storyboard to video' would lift the weaker dimensions.

Suggestions

Add an explicit trigger clause, e.g. 'Use when the user wants an action/fight scene, cinematic combat choreography, or a storyboard-to-video sequence.'

Include natural synonyms users say — 'fight scene', 'action sequence', 'combat choreography', 'storyboard to video', 'cut density' — alongside the model names.

Lead with the plain-language outcome ('Generate a cinematic fight scene video') before the technical pipeline detail, so the 'what' is graspable without model knowledge.

Dimension	Reasoning	Score
Specificity	The description names multiple concrete actions (composing a 16-cell storyboard image, driving Seedance 2.0 image-to-video) and the specific models stacked (GPT-Image-2, Nano-Banana-2, Seedance 2.0 i2v), with only minor gaps in coverage such as the character-sheet and environment phases.	4 / 5
Completeness	The 'what' is clear (generate a fight scene via a storyboard-then-i2v pipeline), but there is no explicit 'when' trigger guidance at all — no 'Use when...' clause or equivalent — which per the guidelines caps completeness and leaves it noticeably below the midpoint.	2 / 5
Trigger Term Quality	It surfaces relevant terms like 'action / fight scene', 'storyboard', and 'image-to-video', but it lacks the natural synonyms a user would say (e.g. 'combat', 'choreography', 'action sequence') and has no 'Use when...' phrasing, missing several common variations.	3 / 5
Distinctiveness Conflict Risk	The narrow niche of storyboard-driven fight-scene generation with named models is clearly distinct from most skills; minor overlap risk exists only with generic image/video-generation skills, keeping it above the midpoint but not at the fully-distinct anchor.	4 / 5
	Total	13 / 20 Passed

Validation

93%

Warnings & errors only

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

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: SamurAIGPT/Generative-Media-Skills
Path: library/motion/ai-fight-scene/SKILL.md
Commit: 5519622

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

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.