CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

seedance-20

This skill should be used when creating, improving, or troubleshooting Seedance 2.0 video on any surface - Dreamina, Jimeng, CapCut, Doubao, Volcengine/Ark, BytePlus, Runway's Seedance route, fal, or third-party provider/router surfaces such as EvoLink, OpenRouter, Kie.ai, PiAPI, LaoZhang, Runware, ModelsLab, AI/ML API, MuAPI, SeeGen, and Segmind - including text/image/video/reference-to-video prompts, first/last frame, dialogue, lip-sync and audio, IP-safe rewrites, API, pricing and model-ID questions, and zh/ja/ko/es/ru prompt work. Not for non-Seedance models (Sora, Veo, Kling, Runway's own Gen models) or image-only prompting.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Failed to scan

The risk profile of this skill

SKILL.md
Quality
Evals
Security

Quality

Content

70%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 well-architected router skill: explicit gate ordering, validation checkpoints, repair loops, and a situation-to-file Load Map, with almost no generic padding in its operational rules. Its weaknesses are verbosity (a philosophy section, a thrice-stated gate, and run-on step paragraphs) and navigation integrity — 24 of 70 links target a skills/ tree that is absent from this bundle, leaving the fast lane's first instruction pointing nowhere.

Suggestions

Consolidate the Director's Read gating into one authoritative statement (the dedicated 'Director\u2019s Read Gate' section) and have the Fast Lane and Loop step 10 reference it in one line each, cutting roughly a third of the repeated doctrine.

Resolve the skills/seedance-*/SKILL.md links against the shipped structure — either restore the sub-skill files or repoint the 24 links to their actual locations (e.g., references/migrated/) so the Load Map and Fast Lane routes resolve from the installed bundle.

Split step 12's mega-paragraph into the quality-pass checklist, shot-table rule, and direct-for-the-model rules as separate bullets, and move the 'Soul' principles into a single compact line each or into a reference file.

DimensionReasoningScore

Conciseness

Mostly efficient — nearly all content is non-inferable Seedance doctrine (gate order, authority tiers, surface-conditional rules) rather than concepts Claude already knows — but there is real tightening headroom: the philosophical "Soul" section ("a person who arrives with a feeling leaves with a film"), the Director's Read gate restated three times (Fast Lane step 2, its own section, and Loop step 10), and step 12's ~400-word single paragraph packing quality pass, shot-table, direct-for-the-model, and moderation rules together. Not 4: these are several genuinely padded or redundant passages, more than 'minor instances of over-explanation'; not 2: what's verbose is skill-specific rules, not explanations of things Claude already knows.

3 / 5

Actionability

Mostly executable guidance for an instruction-only skill: a Load Map naming exact files per situation, ordered gates with jump conditions ("jump straight to the safety gate (step 9)"), hard numeric budgets ("about 40-110 words"), and testable rules ("never write an involuntary outcome as the endpoint… write the deliberate action or leave the consequence off screen"). Not 5: a share of the doctrine stays abstract with no inline worked example — e.g., no sample brief or compact prompt showing the 40-110-word target or a completed two-line non-narrative read — leaving minor gaps a reader must fill from external references.

4 / 5

Workflow Clarity

Clear sequence with explicit validation and feedback loops: the 13-step Operating Loop places checkpoints (source gate "before platform claims", capability check before drafting, quality pass before delivery, moderation pre-screen that "runs last, on the final wording"), error recovery (step 13 retake protocol with verdicts and a troubleshoot route), and checklists (Authority Order for contradictions, sequence invariants). The Fast Lane is explicitly framed as the default path with the full loop one signal away. Not 4: validation steps, repair loops, and conflict-resolution ordering are all explicit rather than minor-gap.

5 / 5

Progressive Disclosure

Scored against the actual bundle: SKILL.md links 70 unique paths; all 46 references/*.md targets exist and are one level deep with a well-signaled Load Map router, but 24 links point to skills/seedance-*/SKILL.md targets that do not exist in this bundle (no skills/ directory; sub-skill bodies appear folded into references/migrated/), and reference files themselves chain onward to sibling references and ../skills/ paths, creating 2+ level navigation for some routes. Not 4: a third of the routing links being unresolvable in the shipped structure is more than a 'minor organization gap'; not 2: the root is a genuine overview with a clear, mostly working navigation index rather than inlined bulk content.

3 / 5

Total

15

/

20

Passed

Description

100%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 model trigger description: explicit 'use when' clause, concrete capability enumeration, comprehensive natural trigger terms across every hosting surface, and explicit negative scope that fences off adjacent models. Dense but not padded — every enumerated surface is a genuine trigger term.

DimensionReasoningScore

Specificity

Lists multiple concrete actions — "creating, improving, or troubleshooting Seedance 2.0 video", "text/image/video/reference-to-video prompts, first/last frame, dialogue, lip-sync and audio, IP-safe rewrites, API, pricing and model-ID questions" — covering the domain comprehensively. Not 4: no minor coverage gaps are evident; every clause names a distinct concrete capability rather than generic language.

5 / 5

Completeness

Explicitly answers both: what — creating/improving/troubleshooting Seedance 2.0 video including prompt types, audio, IP-safe rewrites, and platform-fact questions; when — "This skill should be used when…" with concrete trigger conditions, plus explicit negative scope ("Not for non-Seedance models (Sora, Veo, Kling, Runway's own Gen models) or image-only prompting"). Not 4: the 'when' is fully explicit rather than improvable; not 3: the trigger clause is present, so the cap does not apply. Voice is third person ("This skill should be used"), so no voice penalty.

5 / 5

Trigger Term Quality

Comprehensive natural-term coverage: the surfaces users would actually name ("Dreamina, Jimeng, CapCut, Doubao, Volcengine/Ark, BytePlus, Runway's Seedance route, fal", plus routers like "OpenRouter, Kie.ai, PiAPI"), task words ("prompts", "first/last frame", "lip-sync", "pricing and model-ID"), and language codes ("zh/ja/ko/es/ru prompt work"). Not 4: these are the exact phrases a user on any of these surfaces would say, with synonyms across provider names rather than a few missing terms.

5 / 5

Distinctiveness Conflict Risk

Clear niche with distinct triggers: scoped to Seedance 2.0 across named surfaces, and explicitly excludes the nearest confusion candidates ("Sora, Veo, Kling, Runway's own Gen models", image-only prompting). Not 4: the named-model exclusions remove even the minor overlap risk a closely related skill would carry.

5 / 5

Total

20

/

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

frontmatter_unknown_keys

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

Warning

relative_links

Relative link issues: 38 missing, 42 deeper-than-1-level

Warning

referenced_paths_exist

Referenced path issues: 8 deeper-than-1-level

Warning

Total

13

/

16

Passed

Repository

Emily2040/seedance-2.0

Path

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.