Sports Analytic Skills v0.12.0

This release makes all 24 sports analytics and modeling skills independently usable after a skill-only install. The sports_ds Python toolkit, CLI, loaders, and example pipelines remain optional and are connected explicitly through the sports-ds-bridge skill.

Highlights:

• standalone artifact contracts and helpers across analysis, modeling, validation, simulation, visualization, and reporting
• install-one, install-all, and optional toolkit setup documentation
• aligned 0.12.0 versions across the plugin manifest, toolkit, changelog, and skill metadata
• offline CI on Python 3.10 through 3.13 plus scheduled/manual live-provider CI
• release metadata, skill validation, isolation, and warnings-as-errors gates
• internal planning and exploratory probe artifacts removed from the public tree

Verification:

• 48 offline tests passed on each supported Python version in GitHub Actions
• 7 live provider integrations passed in a fresh hosted Ubuntu environment
• all 24 skills pass the current Codex validator
• sports_ds-0.12.0 wheel builds successfully