seedance-2-0

Generate cinematic clips with ByteDance Seedance 2.0 — the preferred premium video model in OpenMontage when a paid gateway is configured. Use when: (1) producing trailers, teasers, hype edits, or premium cinematic clips, (2) needing native synchronized audio (speech, SFX, ambience) in a single pass, (3) needing multi-shot cuts inside one generation, (4) needing director-level camera control, (5) needing lip-sync from quoted dialogue in the prompt, (6) needing reference-conditioned generation with up to 9 images + 3 video clips + 3 audio clips, (7) wanting consistent character identity across shots. Accessible via fal.ai (`seedance_video` tool), HeyGen (Video Agent / Avatar Shots), Replicate, Runway (Enterprise, non-US), Freepik, BytePlus ModelArk, Higgsfield, Pollo, and other aggregators.

Quality

88%

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

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 executable code, verbatim templates, and a verification checklist, but it is monolithic and over-long: it duplicates prompt-structure guidance and buries time-sensitive pricing/dates inline rather than offloading detail to reference files.

Suggestions

Move time-sensitive details (Feb/April 2026 release dates, Elo ranking, per-second pricing) into a separate '## Model details / pricing' reference file or a clearly-marked dated section so SKILL.md stays lean and staleness is localized.

Consolidate the duplicate prompt-structure methodologies — keep one canonical structure (the Higgsfield Methodology) inline and move the 'Legacy 8-part template' to a reference, or cross-reference it instead of restating both in full.

Offload the opener-template library and the provider-surface table into reference files (e.g. references/prompt_templates.md, references/providers.md) so SKILL.md acts as an overview with one-level-deep, clearly-signaled navigation.

Dimension	Reasoning	Score
Conciseness	The body is dense and assumes Claude's competence (no 'what is a video' padding), but it is long, restates prompt structure twice (Higgsfield Methodology plus a 'Legacy 8-part template'), and embeds time-sensitive data inline ('released Feb 2026', 'Elo 1269 (#1 as of Feb 2026)', '$0.3034 / s') outside any deprecated/old-patterns section, which the guideline penalizes.	2 / 3
Actionability	Provides fully executable Python registry/selector snippets, copy-paste-ready verbatim opener templates, explicit fal.ai model IDs, concrete parameter guidance, and a precise 'What to avoid' table — copy-paste ready guidance throughout.	3 / 3
Workflow Clarity	The 'Iteration strategy' is a clear 5-step sequence (block out shape → lock seed → upgrade to standard → extend → keep README), backed by a 'Verification checklist for every Seedance shot' and an explicit anti-drift feedback loop (drop duration, tighten language, cull references), giving explicit validation checkpoints.	3 / 3
Progressive Disclosure	No bundle files exist (references/scripts/assets absent) and the ~280-line body keeps everything inline — the prompt-template library, provider-surface table, and model IDs are content that should be split into one-level-deep reference files rather than packed into SKILL.md, fitting the anchor where 'content that should be separate is inline.'	2 / 3
	Total	10 / 12 Passed

Description

100%

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: third-person, concrete, with an explicit enumerated 'Use when' clause and distinct, capability-level triggers that separate it from sibling video skills.

Dimension	Reasoning	Score
Specificity	Lists many concrete actions — 'producing trailers, teasers, hype edits', 'native synchronized audio (speech, SFX, ambience)', 'multi-shot cuts inside one generation', 'lip-sync from quoted dialogue', 'reference-conditioned generation with up to 9 images + 3 video clips + 3 audio clips' — matching the top anchor for multiple specific concrete actions.	3 / 3
Completeness	Explicitly states what ('Generate cinematic clips with ByteDance Seedance 2.0') and when via a seven-item enumerated 'Use when: (1)...(7)' clause, satisfying both halves with explicit triggers.	3 / 3
Trigger Term Quality	Surfaces natural terms a user would say when requesting premium video — 'trailers', 'teasers', 'cinematic clips', 'lip-sync', 'multi-shot cuts', 'director-level camera control' — alongside provider names (fal.ai, Replicate, Runway, HeyGen, Higgsfield) giving broad keyword coverage.	3 / 3
Distinctiveness Conflict Risk	The niche is sharply defined by distinctive capabilities (native synced audio, multi-shot cuts, lip-sync, 12-asset reference conditioning) plus a named model, making accidental triggering of a generic video skill unlikely.	3 / 3
	Total	12 / 12 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning

	Total	13 / 16 Passed

Repository: calesthio/OpenMontage
Path: .agents/skills/seedance-2-0/SKILL.md
Commit: 0af32ce

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.