Source intelligence overview¶
After browser analysis, a11yst can enrich findings with:
| Capability | Purpose |
|---|---|
| Source mapping | Rank candidate locations in your repository |
| Recommendations | Suggest remediation guidance aligned with rules and framework context |
Both are best-effort. They help developers act on findings; they do not modify code automatically.
Pipeline¶
Browser finding (rule, target, evidence)
↓
Repository source index + framework catalog
↓
Framework mapper (html, react, next, vue, nuxt, angular)
↓
Optional ranking when multiple candidates match
↓
Optional recommendation recipes
Configuration¶
sourceAnalysis: {
enabled: true, // default true
ranking: true, // default true
recommendations: true, // default true
},
Disable when you only need raw findings:
sourceAnalysis: { enabled: false },
Public wording¶
Human-facing output uses:
- likely source / probable source — not “exact line guaranteed”
- a11yst severity labels — not provider-specific impact names
JSON may include technical provenance (for example sourceImpact) for automation consumers.