Back to skill
Skillv1.0.0

ClawScan security

page-cro · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignMar 11, 2026, 3:24 PM
Verdict
benign
Confidence
high
Model
gpt-5-mini
Summary
The skill's code and instructions align with its stated purpose (analyzing marketing pages for CRO), request no unrelated secrets, and have no surprising install steps — it's internally coherent.
Guidance
This skill appears to do what it says: it parses HTML and scores CRO signals. Before running it, consider: (1) the bundled Python script will fetch any URL you provide — avoid giving internal or sensitive endpoints to prevent unintended network access; (2) run the script in a sandbox or review the code if you will point it at private resources; (3) the parser is static (HTML-based) and may miss dynamically-rendered content (JS), so results are best used as heuristics rather than a full audit. No credentials are requested and there are no hidden downloads.

Review Dimensions

Purpose & Capability
okName/description match the included assets: SKILL.md provides CRO guidance and the included Python script performs an HTML CRO audit (parsing buttons, forms, trust signals, etc.). No unrelated credentials, binaries, or config paths are requested.
Instruction Scope
noteSKILL.md asks the agent to read a local context file (.claude/product-marketing-context.md) if present (reasonable for context). The included script can fetch a URL (urllib.request) or read a local file for analysis — this is necessary for its purpose but means the agent will make outbound HTTP requests when asked to audit remote pages.
Install Mechanism
okNo install specification; this is an instruction-only skill with a bundled Python script. Nothing is downloaded from external URLs or installed automatically.
Credentials
okThe skill declares no environment variables, no credentials, and no config paths. The resources it accesses (page HTML, optional local context file) are appropriate for CRO analysis.
Persistence & Privilege
okalways is false and model invocation is normal. The skill does not request permanent presence or special system-wide privileges.