v0.2.0 - Add platform fixtures and benchmark runner

Highlights

• Remove public-repository and maintainer placeholders
• Update GitHub Actions to checkout@v7 and setup-python@v7
• Add Shopify Liquid and WooCommerce PHP regression fixtures
• Add reproducible benchmark runner and JSON/Markdown reports
• Expose benchmark execution in CI

Verification

• pytest: 5 passed
• deterministic evals: 7/7 passed
• benchmark fixtures: 7
• expected findings: 10
• detected findings: 10
• missed findings: 0
• unexpected findings: 0

Implemented through #6, #7, and #8.

v0.1.0 - Initial release

Initial release of commerce-agent-bench.

Includes:

• Portable ecommerce agent skills
• Evidence-first review protocol
• Deterministic regression fixtures
• Reproducible evaluation runner

Verification

• pytest: 5 passed
• deterministic evals: 5/5 passed