create-video-seedance-2-fal

Generate a single 4-15s vertical video clip with ByteDance Seedance 2.0 reference-to-video via fal.ai. Multi-image reference (avatar + product + setting), native lip-synced VO + ambient audio (generate-audio on by default), internal multi-cut handling within one render. Routes through the GooseWorks FAL proxy (bills the Ads agent). The default clip atom for AI-creator UGC ads built on the NB2 + Seedance architecture. Validated on beauty-by-earth/video-01.

Quality

72%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/ads/packs/ugc-video-formats/create-video-seedance-2-fal/SKILL.md

Quality

Content

77%

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 skill body with concrete commands, a complete prompt template, and strong validation/recovery loops. Its weaknesses are conciseness (length plus scattered dated stamps and dead memory pointers) and progressive disclosure (over-inline decision rules and references to missing files).

Suggestions

Move the long Decision Rules (0–0e, 1–5) block into a separate referenced file (e.g. DECISION_RULES.md) and keep only the top rules inline, reducing the SKILL.md footprint.

Verify every inline file reference exists in the bundle — prompt-example.md, DELIVERY_STYLES.md, and the feedback_*.md memories are currently missing — or remove the dead pointers.

Relocate dated validation stamps ('Validated 2026-05-23', '2026-06-08', '2026-07-18') into a single 'Validated findings / deprecated notes' section so time-sensitive detail does not pad the live workflow.

Dimension	Reasoning	Score
Conciseness	The body is dense and high-signal but very long, with scattered time-sensitive stamps ('Validated 2026-05-23', 'validated 2026-06-08', '2026-07-18') left inline rather than in a deprecated section, plus repeated dead 'Memory: feedback_*.md' pointers — efficient in places but could be tightened.	2 / 3
Actionability	Provides a copy-paste bash invocation, a full literal prompt template, a curl-based recovery workflow, and a concrete failure-mode table with specific recovery actions — fully executable guidance.	3 / 3
Workflow Clarity	Sequences the render in 6 numbered script steps, pairs it with an explicit Quality Checks verification block, mandates 'STOP and surface, do not auto-retry' on NSFW, and gives a numbered recovery workflow with checkpoints — clear sequencing with validation feedback loops.	3 / 3
Progressive Disclosure	Sections and a References block exist, but the giant inline Decision Rules (0–0e, 1–5) block is content that should be split out, and most inline references (prompt-example.md, DELIVERY_STYLES.md, feedback_*.md) point to files absent from the bundle, so references are not cleanly signaled against the actual structure.	2 / 3
	Total	10 / 12 Passed

Description

67%

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, well-differentiated description that names concrete capabilities and a clear niche, but it lacks an explicit 'Use when…' trigger clause and leans on endpoint jargon that weakens natural trigger terms. Adding a user-facing trigger sentence would lift completeness and trigger_term_quality.

Suggestions

Append an explicit 'Use when…' trigger sentence naming the user-facing situations (e.g. 'Use when generating a short AI-creator UGC ad clip that needs a talking-head avatar, lip-synced VO, and a locked product/identity from reference images').

Soften endpoint jargon in the trigger portion — keep 'Seedance 2.0' and 'fal.ai' but move 'reference-to-video' and 'NB2 + Seedance architecture' out of the trigger phrasing so natural keywords dominate.

Cover common user phrasings ('UGC ad video', 'talking-head product review', 'AI influencer clip') so the description matches how a user would actually ask.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'Generate a single 4-15s vertical video clip', 'Multi-image reference (avatar + product + setting)', 'native lip-synced VO + ambient audio', 'internal multi-cut handling within one render' — matching the score-3 anchor.	3 / 3
Completeness	Clearly states what the skill does, but 'when' is only implied via 'The default clip atom for AI-creator UGC ads' with no explicit 'Use when…' trigger clause, which caps completeness at 2 per the guidelines.	2 / 3
Trigger Term Quality	Mixes natural user terms ('vertical video clip', 'UGC ads', 'lip-synced VO') with endpoint jargon ('reference-to-video', 'NB2 + Seedance architecture'), so it has relevant keywords but is not clean of technical jargon.	2 / 3
Distinctiveness Conflict Risk	The niche is sharply defined by 'ByteDance Seedance 2.0 reference-to-video via fal.ai' and 'AI-creator UGC ads built on the NB2 + Seedance architecture', making it unlikely to fire for the wrong skill.	3 / 3
	Total	10 / 12 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: gooseworks-ai/goose-skills
Path: skills/ads/packs/ugc-video-formats/create-video-seedance-2-fal/SKILL.md
Commit: 8866b2a

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.