Battle-Tested Agent
19 production-hardened patterns for AI agents. Every one earned from failure.
Use this skill when you are:
- hardening an agent that will run repeatedly or autonomously
- tightening memory, verification, or anti-hallucination behavior
- reducing compaction failures, weak handoffs, or orchestration drift
- reviewing an agent workspace for missing production patterns
- debugging why an agent keeps losing context, guessing, or dropping work
Do not use this skill for:
[Description truncada. Veja o README completo no GitHub.]