Output and CI

Terminal output

The text report groups findings per file with rule id, selector and remedy. It uses colour in an interactive terminal and plain ASCII when piped or in CI. With 20 or more findings a summary of the most frequent rules comes first. The showcase shows real runs.

Report files

postAudit({
  reports: {
    json: 'audit-report.json',   // one entry per finding
    markdown: 'audit-summary.md', // table for CI artifacts or PR comments
    sarif: 'audit.sarif',         // SARIF 2.1.0 for GitHub Code Scanning
    html: 'audit-report.html',    // standalone HTML report
  },
});

All four can be written in the same run. The legacy output option writes JSON only.

The finding model

Since 0.7.0 a finding carries two axes: outcome says how certain the statement is (fail, review, pass, untested), severity how heavy the problem is (low … critical). The JSON report also lists rule_runs. Rules the tool cannot serve from static HTML, such as colour contrast, appear there as capability_missing.

Accessibility and document rule ids come from a11y-rules, so a finding has the same id in astro-post-audit, auditmysite and LiveAudit, for example images/alt-missing.

GitHub Code Scanning

- name: Build
  run: npm run build

- name: Upload SARIF
  if: always()
  uses: github/codeql-action/upload-sarif@v3
  with:
    sarif_file: audit.sarif

Diagnostics

Both diagnostics write to stderr, so report output on stdout stays clean.

• progress: a single-line progress bar over the check phases. On by default in an interactive terminal, silent in CI. 'verbose' prints each check as its own line with finding count and timing instead.
• debug: true: resolved configuration, discovery counts and per-check finding counts and timings. Replaces the progress bar.
• benchmark: true: per-check timing breakdown.