redbox-video-director

Use when generating short videos with RedBox official video API. Produces a detailed shot script first, asks the user to confirm it, then chooses between text-to-video, reference-guided, and first-last-frame modes and calls the correct wan2.7 video model with prompt discipline focused on motion, reference elements, and transitions.

Quality

80%

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 ./desktop/electron/builtin-skills/redbox-video-director/SKILL.md

Quality

Content

85%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 highly actionable, well-sequenced procedural skill with concrete commands, model mappings, and explicit confirmation/stability checkpoints gating costly generation steps. Its main weakness is progressive disclosure: at ~240 lines with no reference files, some detailed material could be split out for a leaner overview.

Suggestions

Extract the Prompt Consistency Rules / Keyframe Generation Order and the Script Format spec into one-level-deep reference files (e.g. references/keyframes.md, references/script-format.md) and link from the body, so SKILL.md stays a concise overview.

De-duplicate the storyboard-first / confirmation guidance that recurs in Default Workflow, Production Strategy, and Storyboard-First Rule so each constraint is stated once.

Tighten verbose bullets like the subject-library reuse note ('Do not keep reusing raw subject-library portraits... unless you truly need extra补充 angles or missing objects') to reduce token cost.

Dimension	Reasoning	Score
Conciseness	Dense, almost entirely procedural rules with no padding and no over-explanation of concepts Claude already knows; held back from 5 by minor restating of the storyboard-first / confirmation guidance across the Workflow, Production Strategy, and Storyboard-First sections.	4 / 5
Actionability	Fully executable guidance: exact app_cli(command="...") invocations, model mappings (wan2.7-t2v/r2v/i2v-video), concrete image-count limits (0, 1-5, exactly two in 首帧/尾帧 order), a markdown table template, and labeled prompt format ('Image 1: ...') covering all three modes.	5 / 5
Workflow Clarity	A clear 10-step sequence with explicit validation checkpoints ('If the user has not yet confirmed the script, do not generate the video'), storyboard-first gating, and a keyframe-stability feedback loop ('Only after the keyframes are stable should video generation proceed').	5 / 5
Progressive Disclosure	No bundle files exist and the single ~240-line file is well-sectioned, but the detailed keyframe prompt-consistency rules and script-format spec are long enough that they could plausibly live in one-level-deep reference files, leaving the body a tighter overview.	3 / 5
	Total	17 / 20 Passed

Description

75%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.

A specific, third-person description that states concrete actions and an explicit 'Use when' trigger scoped to the RedBox video API. It is just shy of perfect mainly because the trigger-term coverage could include a few more natural user phrasings.

Suggestions

Add a few everyday user phrasings to the trigger (e.g. 'make a video', 'animate this image', 'turn images into a video clip') to improve trigger-term quality.

Mention the project-pack / multi-clip concatenation capability briefly so the 'what' covers the full scope of the skill.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions ('Produces a detailed shot script first, asks the user to confirm it, then chooses between text-to-video, reference-guided, and first-last-frame modes and calls the correct wan2.7 video model'), with only minor gaps such as omitting project-pack/concatenation work.	4 / 5
Completeness	Both 'what' (shot script, mode selection, model call, prompt discipline) and 'when' ('Use when generating short videos with RedBox official video API') are present; the trigger could be slightly more explicit with concrete user phrases.	4 / 5
Trigger Term Quality	Good natural-term coverage ('short videos', 'image-to-video', 'reference-image video', 'first/last frame transitions') plus specific mode names, but a few common user phrasings like 'make a video' or file extensions are absent.	4 / 5
Distinctiveness Conflict Risk	Clear niche scoped to the RedBox official video API with named wan2.7 models and three explicit modes, giving minimal conflict risk with unrelated skills.	5 / 5
	Total	17 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: Jamailar/Beav
Path: desktop/electron/builtin-skills/redbox-video-director/SKILL.md
Commit: e8cbd6a

Reviewed: 4 days 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.