#ai
2 posts-
Building an AI Agent Layer for E2E Test Automation
How I structured a knowledge-driven AI agent setup using Cline that lets any developer generate, debug, and maintain E2E tests with minimal domain knowledge — by encoding team patterns into machine-readable files.
-
Local AI Code Reviews on GitHub PRs with LM Studio and GitHub Actions
How I set up automated PR code reviews using a locally running Qwen2.5 Coder model via LM Studio, Tailscale, and a self-hosted GitHub Actions runner — with a graceful fallback when the PC is off.