BrowserAgentRegression v1.1.2

BrowserAgentRegression v1.1.2 is a bounded software maintenance patch. The immutable v1.0.0 protocol and evidence baseline remain unchanged.

Fixed

• Gave all three checkpoint nodes in each Studio row equal geometry.
• Reserved a dedicated annotation track and added safe wrap opportunities for long checkpoint IDs.
• Added real-browser regression checks for containment, equal height, aligned tracks and percentage/annotation overlap across five desktop widths.
• Updated current software identity and quick-start refs to v1.1.2.

No task, variant, driver, checkpoint, fixture, schema, protocol, dependency, evidence or research claim changed.

Provenance

• Merge commit: 7738e5f8b522a183aaf5aa07343095533af9a403
• Pull request: #12
• Exact-main CI: 33376654170

Distribution SHA-256

• browser_agent_regression-1.1.2-py3-none-any.whl: 6503ee6de5987d140d78c5c77eaa486a97cd9c686d2a5cdb3ce05bed42de290b
• browser_agent_regression-1.1.2.tar.gz: 042055bf71e432233a3152199bb0a07cba63461e2661486e51b52a7bdb31c0dc

Both artifacts were installed in separate fresh Python 3.11 environments; package identity, dependency checks, CLI doctor, frozen-evidence verification and packaged Studio server construction passed.

BrowserAgentRegression v1.1.1

v1.1.1 is a documentation correction over the v1.1.0 presentation and maintenance closure; the immutable v1.0.0 research protocol/evidence baseline, results, behavior, dependencies, and claims are unchanged.

This patch corrects the current-software Quick Start to clone the release that contains Studio and aligns package, CLI, Studio, changelog, and maintenance identity with v1.1.1. The uploaded wheel and source distribution were built from main commit 8deaf8189b319d1ec7b8f527d763892a62c9ad97 and installed in fresh Python 3.11 environments.

v1.1.0 — Presentation and Maintenance Closure

v1.1.0 is the presentation and maintenance closure release built on the unchanged v1.0.0 research/evidence baseline.

This release finalizes the Studio checkpoint-divergence presentation, shared AlvenX interface, responsive and accessible interaction, maintenance documentation, and installable Python packaging. The frozen v1.0.0 protocol, fixtures, evidence, and claims are unchanged.

SHA-256:

• browser_agent_regression-1.1.0-py3-none-any.whl: 461c314cd8336d678f89e645f74ef34dab03c40ce8238913435938692ea131f3
• browser_agent_regression-1.1.0.tar.gz: 067b0d057b65cb9f53d2d4d3d2fbcd32cbd02edf206b9dc20227670f9b662e1b

BrowserAgentRegression v1.0.0

First stable technical release of the local-first browser-agent regression harness.

Highlights:

• Frozen v1 evidence schema and controlled-UI protocol.
• Deterministic Chromium oracle and calibrated failure variants.
• doctor and semantic verify commands with actionable exit codes.
• Committed, hash-checked v1 evidence: 36/36 oracle passes and three 100%→0% calibrated regressions.
• Python 3.11–3.13 CI plus clean-wheel smoke tests against a real Chromium runtime.

External adoption remains unvalidated and is not claimed by this release.