PDCA memory loop for Claude Code — prevents "AI fixed bug X and broke feature Y" across iterations. Installs a three-tier memory system in your project (project-memory, decisions-log, task-history) and enforces a closed PDCA loop on every task: Plan (read memory before coding), Do, Check, Act (write memory back after). The Act step is what most tools skip — without it, your context file rots after 10 iterations. With it, iteration 50 has the same context quality as iteration 1. Complements superpowers: superpowers brings per-session discipline, ai-coding-ok brings cross-session memory. Also works with Copilot, Cursor, and OpenCode via the same templates.