agentsreview
Reviewing AI-generated code with multiple models
How I use subagents, the Codex and Claude Code CLIs, and a repair loop to review changes before accepting them.

Independent notes on software development, AI agents, and engineering practice.
How I use subagents, the Codex and Claude Code CLIs, and a repair loop to review changes before accepting them.

How I turn a high-level request into a reviewed specification, delegate the work, and decide when to inspect the final change myself.
