This is a comprehensive framework for profiling and optimizing systems that use multiple AI agents working together. It tackles the real pain points: identifying bottlenecks across agent workflows, reducing unnecessary context window bloat, minimizing token costs through smart model selection, and coordinating parallel execution without agents stepping on each other. The source includes actual orchestration code for distributing work across database, application, and frontend performance agents. Best suited for teams already running multi-agent systems at scale where latency, throughput, or LLM costs have become measurable problems. If you're just tuning a single agent or haven't instrumented anything yet, this is overkill.
npx skills add https://github.com/sickn33/antigravity-awesome-skills --skill agent-orchestration-multi-agent-optimize