Back to skill
Skillv1.0.0
ClawScan security
nielsen-ui-review · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 18, 2026, 4:06 PM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- This is an instruction-only UI review skill whose required behavior, inputs, and permissions align with its Nielsen-heuristics review purpose and it requests no extra environment access or installs.
- Guidance
- This skill appears coherent and low-risk, but be mindful when sharing screenshots: redact or remove passwords, tokens, user PII, or other sensitive data before upload. Confirm where images will be processed/stored (local vs cloud) according to your agent/platform privacy settings. When you request a review, answer the short clarification questions or accept the skill's stated assumptions so findings are appropriately targeted.
Review Dimensions
- Purpose & Capability
- okThe name/description promise a heuristic UI review and the SKILL.md contains detailed step-by-step guidance for performing that review. There are no unrelated requirements (no credentials, binaries, or config paths) that would be disproportionate to a UI review task.
- Instruction Scope
- okInstructions are narrowly scoped to inspecting a provided screenshot, asking clarifying questions, and producing findings mapped to Nielsen's heuristics. The skill does not instruct the agent to read unrelated files, access environment variables, or transmit data to unexpected endpoints. It does advise making explicit assumptions if the user doesn’t respond.
- Install Mechanism
- okNo install spec or code files are present; this is instruction-only so nothing is written to disk or fetched during install. That minimizes installation risk.
- Credentials
- okThe skill declares no required environment variables, credentials, or config paths. The only input expected is user-provided screenshots and minimal context questions, which is proportionate to the stated purpose. Note: screenshots can contain sensitive data—see user guidance.
- Persistence & Privilege
- okalways is false and the skill does not request persistent system-wide privileges or config changes. disable-model-invocation is false (normal), so the agent could invoke the skill autonomously per platform defaults, but that is not unusual for a user-invocable skill and is not combined with other red flags here.