Run the automated scanner on a sample and review its proposed findings.
Goal
Run an automated axe-core scan on a sample and merge the proposals into the audit's findings.
Preconditions
- The workspace has the Scanner add-on activated, or is on Enterprise.
- The sample's URL is publicly reachable, or the workspace has a BYOK Browserless key with appropriate access.
- The sample is part of an active audit.
Steps
- Open the sample in the audit's Samples tab.
- Click Run scan (or the equivalent on the sample row).
- Wait for the scan to complete (typically under 30 seconds per sample).
- The Scan Review modal opens, showing axe-core violations mapped to criteria in the audit's referential.
- For each proposed criterion, decide:
- Accept — apply the verdict (typically Failure) with the suggested problem and solution.
- Skip — do not apply.
- Edit then accept — adjust problem and solution before applying.
- Click Apply selected. Only "Not tested" criteria are overwritten; manual decisions are preserved.
Batch scanning
For multiple samples at once, use Batch scan from the Samples tab. The batch aborts after 2 consecutive failures to avoid wasted runs.
Verify success
- Accepted proposals appear as findings on the relevant samples.
- Findings have
scanner_source: trueset (visible in the finding metadata). - Manual decisions are unchanged.
Common pitfalls
- Trusting the scanner blindly. Axe-core catches around 30 to 50 percent of WCAG failures. The auditor still evaluates the rest manually.
- Leaving scanner-suggested text unedited. Auto-generated descriptions often need rewriting for clarity. See methodology / finding quality.
- Scanning behind authentication without a BYOK Browserless key configured for it. The shared scanner can only see public pages.
Related
- product / integrations
/docs/future-browserless-replacement.md— planned in-house replacement.