CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

seedance-sequence

This skill should be used when a Seedance 2.0 request is a long story, connected set of clips, multi-generation scene, campaign sequence, dense storyboard, continuation-ready plan, or any idea that must be divided into stateful clips.

Quality

73%

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

Fix and improve this skill with Tessl

tessl review fix ./skills/seedance-sequence/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

77%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 body is an efficient, well-sequenced operational spec with explicit validation gates and concrete state schemas — workflow clarity is a genuine strength. Its main defects are delegated substance living in reference files that are missing from the evaluated bundle, and one overloaded step-2 paragraph that hurts readability more than clarity of sequence.

Suggestions

Ship or fix the reference bundle: ../../references/directors-read.md, prompt-compiler.md and the other eight linked files are absent, so the skill's core mechanisms cannot be loaded; verify the relative paths resolve from the skill directory.

Split Build Process step 2 into separate steps (lane classification, ten-field record, directing-engine spine) — the current single sentence buries five instructions.

Deduplicate the reference-load instructions between the opening paragraph and the Build Process, or state the loading rule once and reference it.

DimensionReasoningScore

Conciseness

The body is dense and assumes competence — no space is spent explaining what video generation or continuity is; every section carries operational rules. Minor trimming is possible: step 2 of the Build Process is a single ~120-word sentence packing five distinct instructions, and the reference-load list in the opening paragraph repeats content stated again in the Build Process, keeping it just below the 'lean, every token earns its place' anchor.

4 / 5

Actionability

For an instruction-only skill the guidance is highly concrete: exact field lists for clip and scene cards, specific defaults ('max_chain_depth (default 2, hard ceiling 3)'), a worked micro-example ('exit terminal and reach open car door'), and a numbered 11-item output contract. Below 5 because several key mechanisms (the Director's Read record, the prompt compiler, the transfer contract) are delegated to reference files that are not present in the bundle, leaving those steps not fully executable as written.

4 / 5

Workflow Clarity

The 11-step Build Process is explicitly ordered with real validation checkpoints: 'After generation, require the clip or final frame, record observed start/end state, reconcile physical and dramatic handoffs, and only then compile the next prompt', plus the intent-echo gate 'confirmed before generation spends money' and the chain-cap/re-anchor rules. This matches the anchor of clear sequence with explicit validation and a feedback loop; the only friction is readability of the step-2 run-on sentence, not the sequence itself.

5 / 5

Progressive Disclosure

Structurally the body is a good overview with ten clearly named, one-level-deep, conditionally loaded references ('Load dense-storyboard-mode when the request contains many shots'), which reads like a 4-5. However, none of the referenced files (../../references/directors-read.md and the other nine) exist in the bundle — no references/ directory is present — so the disclosure structure is not actually navigable and key delegated content is inaccessible. That drops it to 'references present but not resolvable / organization incomplete'.

3 / 5

Total

16

/

20

Passed

Description

61%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 a strong, explicit trigger clause with good natural vocabulary, but it is one-sided: it says when to use the skill and almost nothing about what the skill does. Adding one sentence of concrete capabilities (global story planning, per-clip prompt compilation, continuity tracking) would lift both completeness and specificity.

Suggestions

Add an explicit 'what' sentence, e.g. 'Plans the complete story as stateful scenes and clips, and compiles one Seedance clip prompt at a time so each continuation matches accepted footage.'

Include the natural continuation phrases users type ('continue', 'extend', 'next part', 'part two', 'make it longer') in the description, not only in the body.

Clarify the boundary with the parent seedance-20 skill (e.g. 'use for multi-clip sequences, not single standalone clips') to reduce overlap risk.

DimensionReasoningScore

Specificity

The description names the domain ('Seedance 2.0 request') and one concrete action-like phrase ('must be divided into stateful clips'), but the capabilities of the skill itself (planning the story globally, compiling one clip prompt at a time, continuity QC) are never stated. It is above 'Names the domain but actions are minimal' (2) because 'divided into stateful clips' is a concrete operation, but below 4, which expects several specific actions listed.

3 / 5

Completeness

The 'when' is explicit and rich ('This skill should be used when...'), but the 'what' is only weakly implied — the description never states what the skill does (plan the sequence, compile clip prompts, track continuity). Per the anchors, 'only when is present without a clear what' sits at 2, but 'must be divided into stateful clips' conveys a partial what, landing between 2 and 3; the what is too thin for 4.

3 / 5

Trigger Term Quality

Strong natural trigger vocabulary: 'long story', 'connected set of clips', 'multi-generation scene', 'campaign sequence', 'dense storyboard', 'continuation-ready plan'. Below 5 because the common continuation phrases users actually type ('continue', 'extend', 'next part', 'part two', 'make it longer') appear only in the body, not in the description.

4 / 5

Distinctiveness Conflict Risk

A clear niche (Seedance 2.0 multi-clip sequencing) with distinct triggers; minimal conflict risk against unrelated skills. Below 5 because it likely overlaps with the parent 'seedance-20' skill for single-storyboard or multi-scene requests, and 'long story' / 'dense storyboard' could fire for the non-sequence path too.

4 / 5

Total

14

/

20

Passed

Validation

81%

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

CriteriaDescriptionResult

metadata_field

'metadata' should map string keys to string values

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

relative_links

Relative link issues: 14 suspicious

Warning

Total

13

/

16

Passed

Repository

Emily2040/seedance-2.0

Path

skills/seedance-sequence/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.