Skip to content

Classifications

a11yst uses two related concepts:

Concept Meaning
Lifecycle status How a finding relates to the baseline comparison (new, known, regressed, resolved)
Disposition Optional team classification stored in the baseline (false-positive, accepted-risk, …)

Lifecycle status comes from comparison. Disposition comes from a11yst classify and affects CI policy filtering.

Lifecycle statuses

New

The finding’s fingerprint is not in the baseline.

First time you see this issue in the comparison reference.

Known

The finding matches a baseline entry and no regression was detected.

The issue still exists in the application; it is unchanged relative to the baseline snapshot.

Regressed

The finding matches a baseline entry but a regression was detected. Common reasons:

Reason Meaning
returned-after-resolution Was resolved relative to baseline but appeared again
classification-expired An accepted classification expired
severity-increased Severity rose (for example HIGH → CRITICAL)
confidence-increased Confidence rose for a11yst-owned findings
scope-expanded Scope expanded relative to baseline

Resolved

The fingerprint exists in the baseline but is absent from the current audit (within comparable coverage).

Useful for confirming fixes without re-triaging every historical issue.

Example

Baseline (RUN 1)

Severity Rule Status in baseline
HIGH link-name stored
MEDIUM color-contrast stored

After code change (RUN 2)

Finding Lifecycle
link-name (still present) known
color-contrast (fixed) resolved
select-name (first appearance) new
link-name (severity increased) regressed

Terminal and JSON output use these exact status labels.

Dispositions

Assign a disposition when your team needs to record why a known finding is handled differently in CI:

a11yst classify <finding-id> \
  --disposition false-positive \
  --reason "Decorative icon hidden from assistive tech" \
  --yes
Disposition CI policy impact
false-positive Excluded from policy breaches
not-applicable Excluded from policy breaches
accepted-risk May still breach policy if severity meets threshold
third-party May still breach policy
manual-review May still breach policy

Remove a disposition:

a11yst unclassify <finding-id> --yes

Filtering findings

a11yst findings --status new --status regressed
a11yst findings --disposition accepted-risk

Policy interaction

CI policies evaluate new and regressed findings that meet the configured --minimum-severity. Dispositions false-positive and not-applicable are never policy breaches.

See Policies and Exit codes.