Back to skill
Skillv2.0.0
ClawScan security
Seedance 2.0 prompt-engineering skill · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 4, 2026, 12:02 AM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill is internally consistent with its stated purpose (Seedance 2.0 prompt engineering); it is instruction-heavy, requests no credentials, has no network installs, and the included scripts are harmless local scaffolding and examples.
- Guidance
- This appears to be a safe, instruction-only prompt-engineering skill: it asks for no credentials, has no network installs, and the scripts only create local folders or print templates. Before installing, confirm you trust the skill owner (source unknown), avoid pasting private or copyrighted assets into prompts if you don't want them uploaded to external services, and review any generated prompts before using them with Seedance to ensure compliance with your platform and legal requirements.
Review Dimensions
- Purpose & Capability
- okName/description (prompt engineering for Seedance 2.0) match the SKILL.md guidance and included example recipes. There are no unrelated env vars, binaries, or install requests.
- Instruction Scope
- okRuntime instructions are limited to generating structured prompts, asset mapping, timeline beats, moderation/IP avoidance guidance, and example templates. They do not instruct reading local secrets, contacting external endpoints, or accessing unrelated system paths.
- Install Mechanism
- okNo install spec. The only code files are a harmless example test runner (SKILL.sh) and a local workspace setup script; neither downloads or executes remote code or writes outside a user-specified workspace.
- Credentials
- okThe skill requires no environment variables or credentials. The guidance references platform-side asset placeholders (@image1/@video1) which is appropriate for a prompt-engineering skill.
- Persistence & Privilege
- okalways is false and the skill does not request persistent/system-wide changes. The included scripts only create a local directory structure for drafts/inputs/outputs.