v0.1.1 - experimental prerelease

v0.1.1 - experimental prerelease

Release classification: experimental prerelease

Here, prerelease is the GitHub Release classification. The Python package
version is 0.1.1, and the workflow does not publish it to PyPI.

This version is an experimental review aid. Output is not a compliance determination,
an ATO assessment or professional advice. A qualified human must reconcile the source
records, decide the legal and payroll classifications, and approve any payment,
disclosure, lodgment or accounting action.

What changed

• Refreshed the implementation against the legislation and official material
  current on 15 August 2026, including final LCR 2026/1, LCR 2026/2 and LCR
  2026/3. LCR 2026/D1 remains draft.
• Made the July 2026 transition, statutory contribution allocation and item 4
  deadline extension fail closed when the supplied records do not establish the
  facts. Material uncertainty is attention-driving UNKNOWN with no exposure
  figure.
• Required an actual subsequent non-out-of-cycle payment for the out-of-cycle
  pathway, as required by F2026L00784.
• Corrected the whole-of-jurisdiction holiday boundary and made post-horizon
  funded, unfunded and stale-prepayment outcomes fail closed.
• Clarified that regulations 11 and 12, qualifying earnings, termination
  treatment and the maximum contribution base remain operator inputs or review
  boundaries rather than decisions made by the tool.
• Retained experimental labels on SG-charge estimates and removed language that
  could imply a settled liability or compliance result.

Release artefacts and verification

The release workflow builds the wheel and sdist twice from the exact tagged Git
tree in the same locked release job and refuses publication unless the bytes
match. This is a same-job repeatability check, not a cross-platform or
future-toolchain byte-identity claim. It also publishes deterministic UTC/LF
source archives, a deterministic SPDX 2.3 runtime SBOM, the reviewed notes and a
sorted SHA256SUMS file. GitHub build-provenance
attestations cover every uploaded artefact, with a separate SBOM attestation for
the wheel.

After downloading the files, verify them before installation:

tag_sha=$(git ls-remote \
  https://github.com/ryanduguid/payday-super-checker.git \
  'refs/tags/v0.1.1^{}' | cut -f1)
test "${#tag_sha}" -eq 40
sha256sum --check SHA256SUMS
gh release verify-asset v0.1.1 \
  payday_super_checker-0.1.1-py3-none-any.whl \
  --repo ryanduguid/payday-super-checker
gh attestation verify payday_super_checker-0.1.1-py3-none-any.whl \
  --repo ryanduguid/payday-super-checker \
  --source-digest "$tag_sha" \
  --source-ref refs/heads/main \
  --signer-workflow \
    ryanduguid/payday-super-checker/.github/workflows/release.yml
gh attestation verify payday_super_checker-0.1.1-py3-none-any.whl \
  --repo ryanduguid/payday-super-checker \
  --source-digest "$tag_sha" \
  --source-ref refs/heads/main \
  --signer-workflow \
    ryanduguid/payday-super-checker/.github/workflows/release.yml \
  --predicate-type https://spdx.dev/Document/v2.3

The release workflow does not publish to PyPI. Install only the wheel you have
verified, preferably in an isolated environment.

Known review boundaries

• The bundled official holiday horizon ends on 31 August 2027.
• The importer cannot infer fund-receipt order or SG-charge assessment facts;
  its statutory-allocation confirmation records an operator reconciliation.
• Fund receipt and allocation evidence are required before item 4 can extend a
  deadline.
• Raw-pay classification, regulations 11 and 12, LCR 2026/D1 matters, cumulative
  maximum-contribution-base calculations, choice loading, late-payment penalty
  and post-assessment GIC remain outside the automated result.
• The ATO determines the final charge and applies the assessment-level rounding
  rule. Monetary output here remains an experimental estimate.