Audit a page set against a named WCAG version and level, classify issues by severity/level, and compute a weighted conformance score. Reach for this on a conformance question.
Installs into .claude/skills of the current project.
Are you the author of Run Wcag Audit?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/mcorbett51090-run-wcag-audit)
---
name: run-wcag-audit
description: "Audit a page set against a named WCAG version and level, classify issues by severity/level, and compute a weighted conformance score. Reach for this on a conformance question."
---
# Skill: Run WCAG audit
An audit with no named target or that rests on an automated scan alone is not a conformance claim (§3 #1 #2).
## Step 1 — Name the target
WCAG version + level (A/AA/AAA) and the page set in scope — conformance is per-criterion at the target level (§3 #1).
## Step 2 — Scan, then test by hand
Run the automated scan as a first pass, then manually check the human-judgment criteria it cannot detect (§3 #2).
## Step 3 — Classify each issue
By severity and WCAG level; one failing Level-A criterion fails the page (§3 #2).
## Step 4 — Score conformance
Weighted score + critical-blocker flag via `accessibility_calc.py conformance` (§3 #1).
## Output
A criterion-level audit with severity/level classification, critical blockers, and a weighted conformance score. Traverse Tree 1 in the decision-trees file.
