v0.1.0 — provider-independent AI code review

First public GitHub Marketplace release of the AgentsKit Code Review action.

• review pull requests through Codex CLI, Claude CLI, hosted APIs, or Ollama
• run seven review lenses followed by adversarial verification
• post PR findings and optionally fail only above a configured severity
• keep secrets in environment variables

Suggested Marketplace categories: Code quality and Continuous integration.