Back to skill
Skillv1.0.0
ClawScan security
image-prompt-engineer · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMay 1, 2026, 1:45 PM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5.5
- Summary
- This instruction-only skill appears to provide image-generation prompt-writing guidance without requesting credentials, tools, installation steps, persistence, or system access.
- Guidance
- This looks safe to install as a prompt-assistance skill. As with any image-prompting helper, review generated prompts for appropriateness and platform policy compliance before using them.
Review Dimensions
- Purpose & Capability
- okThe artifacts are coherent with the stated purpose: they contain prompt architecture, style, lighting, composition, negative prompt, and platform-optimization reference material for image generators.
- Instruction Scope
- okThe instructions focus on generating and refining text prompts; they do not direct the agent to override user intent, call tools, access accounts, or perform high-impact actions.
- Install Mechanism
- okNo install specification, required binaries, environment variables, credentials, or code files are present.
- Credentials
- okThe skill does not ask for local file access, network access, account access, shell execution, or other environmental permissions.
- Persistence & Privilege
- okNo persistence, background behavior, memory storage, credential use, or privilege escalation is evidenced in the artifacts.