Skillv1.0.0

ClawScan security

WCAG 2.1 AA Web UI Audit · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignMar 4, 2026, 5:03 PM

Verdict: Benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's code, instructions, and included automation script are consistent with an accessibility (WCAG 2.1 AA) audit — nothing in the package requests unrelated credentials or hidden endpoints.
Guidance: This skill appears coherent for performing WCAG 2.1 AA audits. Before installing or running it, consider: (1) Review the run_axe_playwright.js source (included) yourself — it only launches a headless browser and writes local output files. (2) If you plan to run the automation, install playwright and @axe-core/playwright from npm in a controlled environment; be aware Playwright may download browser binaries. (3) Only provide URLs you trust (do not point it at sensitive internal systems unless you intend that scan and have appropriate credentials). (4) The skill's origin/homepage is unknown — if you need stronger assurance, obtain the package from a known repository or request provenance from the publisher. Otherwise the package is internally consistent and aligned with its stated purpose.

Review Dimensions

Purpose & Capability: okThe name/description (WCAG 2.1 AA audit) match the provided artifacts: thorough SKILL.md, WCAG checklists, templates, and an optional automation script using axe + Playwright. There are no unrelated environment variables, binaries, or config paths requested.
Instruction Scope: noteRuntime instructions stay within audit scope (manual checks, automated axe+Playwright runs, templates for findings). Caution: the optional script will visit any URLs you pass to it — including internal/staging sites — so only supply URLs you intend to have scanned and be mindful of authentication/credentials required by those environments (the skill does not request or manage secrets).
Install Mechanism: noteThere is no install spec — the script uses dynamic imports for 'playwright' and '@axe-core/playwright'. Running the automation requires installing those npm packages; Playwright commonly downloads browser binaries when installed or first run. This is expected for a web-scanning tool but is the primary installation-related risk to be aware of (network downloads of browser executables from upstream sources).
Credentials: okThe skill requests no environment variables, credentials, or config paths. Its behavior (visiting user-provided URLs and writing local output files) is proportionate to an accessibility scanner. If you run against protected sites you will need to supply any required auth out-of-band (the skill does not collect secrets).
Persistence & Privilege: okThe skill is not forced-always, does not claim persistent system privileges, and does not modify other skills or global agent settings. The only filesystem writes are the expected local outputs/axe-results.json and outputs/axe-summary.md created by the optional script.