Common Workflows

Real-world workflows for the most common use cases.

A. Local development scan

The simplest workflow: scan your local project during development.

bash

cd ~/projects/my-app
wcag-audit scan

# Opens your WCAG_FIXES.md in Cursor / Claude Code
# Fix issues → re-scan to verify

Use --no-ai during rapid iteration to skip the slower AI vision step. AI vision is included on every paid plan — run a full scan with AI before shipping.

B. Scanning authenticated pages

60-75% of real-world a11y violations live behind login. Use --auth to audit dashboards, admin panels, and role-gated routes.

bash

# Interactive: browser opens, you log in
wcag-audit scan --auth

# Export session for reuse
wcag-audit scan --auth --export-auth ./auth.json

# Reuse in future scans (headless)
wcag-audit scan --auth-storage ./auth.json

C. CI/CD pipeline gate

Block PRs that introduce accessibility regressions. The ci command exits with code 1 when issues exceed your threshold.

GitHub Action

yaml

# .github/workflows/a11y.yml
name: Accessibility
on: [pull_request]

jobs:
  wcag:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - uses: wcagauditdev-commits/wcag-documentation-creator/action@main
        with:
          license-key: ${{ secrets.WCAG_LICENSE_KEY }}
          fail-on: critical

Any CI provider

bash

# Install deps, start dev server, then:
wcag-audit ci --fail-on critical

# With auth:
wcag-audit ci --fail-on serious --auth-storage ./auth.json

D. Auditing a deployed site

No local source needed. Point the CLI at a URL and it crawls via BFS.

bash

# Crawl 2 levels deep (default)
wcag-audit scan --url https://staging.example.com

# Deeper crawl
wcag-audit scan --url https://example.com --crawl-depth 4

E. AI editor integration

Generate output files your AI coding tool can consume directly.

json

// .wcagauditrc
{
  "outputs": [
    "excel",
    "markdown",
    "ai-fix",
    "cursor-rules",
    "agents-md"
  ]
}

Output formats and recommended use cases
Output	Best for
`markdown`	Paste into any AI chat (Claude, GPT, Copilot)
`ai-fix`	Structured JSON for programmatic consumption
`cursor-rules`	Auto-attaches as context when editing matching files in Cursor
`agents-md`	Upserts a section in your AGENTS.md for Claude Code

F. Watch mode during development

Auto-rescan when source files change. Great for fixing issues one by one.

bash

wcag-audit watch

Scan results are cached for 24 hours per route. Only changed routes are re-scanned, saving credits. Use --no-cache to force a full rescan.

Next step: Troubleshooting common issues →