Production-grade design blueprint for stable, effective AI agents — system design, tool design, context/memory manageme…
Production-grade design blueprint for stable, effective AI agents — system design, tool design, context/memory management, pre-action verification, security hardening, observability, human-in-the-loop, and evaluation. Use proactively whenever designing, building, hardening, auditing, debugging, or shipping any agent or agentic system — tool/function definitions, MCP servers, multi-step or autonomous workflows, model routing, agent memory, guardrails, agent failures, voice agents, workflow-automation agents, or agent-based client deliverables. Trigger on agent, agentic, autonomous, tool use, MCP, orchestration, subagent, agent eval/failure/hardening — even without "best practices" asked. Default is passive reference — surface only relevant principles. On "walk me through agent design" or "phase this build," run the phased workflow explicitly.
superlogicai
cli
free
Others in the same category, ranked by how often they are opened.
