Bugbix 1.0.0

2026-10-06

• Establish the v1 CLI, verdict/exit and report schema 2 stability contract for selected Python regression checks.
• Retain the verified unittest/pytest engine, final xfail/xpass handling and original-checkout preservation behavior.
• Deliver a prebuilt wheel and installable source archive with MIT licensing and SHA256 checksums.
• Verify both actual artifacts in fresh environments, including positive/negative verdicts, recovery and checkout preservation.
• Gate main-branch release publication on the existing Windows/Linux and Python 3.11/3.12 checks and complete artifact verification.

Download the prebuilt wheel or source archive and verify SHA256SUMS. Python 3.11+ and Git are required. The runtime has no package dependencies. Distributions include the MIT license; source installation uses the declared build backend. The v1 CLI/report contract and recovery guide describe the supported regression-check scope.

Installation and upgrade instructions.

Bugbix 0.4.0 alpha 2

The pytest runner now reads final test reports after expected-failure handling. Suites containing xfail or xpass stay inconclusive, including a mixed suite that previously reported a regression. Ordinary assertion failures and runtime errors remain distinct.

The package, CLI, and report versions now agree on 0.4.0a2. Install the attached wheel with python -m pip install bugbix-0.4.0a2-py3-none-any.whl, or follow the tagged-source installation in the README.

Validated on Windows/Python 3.12.14 with pytest 9.1.1: 22 product tests, five installed-wheel CLI cases, and a check of Bugbix's own new regression test against the pre-fix revision. See VERIFICATION.md for scope and earlier trials.

Bugbix 0.4.0 alpha 1

Bugbix is a local Python CLI that checks whether selected regression tests fail on a known pre-fix Git commit and pass on a fix. This is the first public alpha.

Install

python -m pip install "git+https://github.com/agammann/bugbix.git@v0.4.0-alpha.1"

You can also install the attached wheel. Bugbix is not on PyPI.

Verified locally

• 20 product tests passed on Windows with Python 3.12.
• The installed alpha wheel observed the expected regression in boltons PR #513 and rejected a passing control test in more-itertools.
• Earlier focused trials reproduced six fail-before/pass-after tests across those two public projects.

Generated build files can prevent test collection; the humanize trial was inconclusive for that reason. Tests run with your user permissions, so use trusted repositories. See Verification and the README for exact scope and instructions.