A skill package that teaches agents context engineering, memory, and evaluation design
A skill package that teaches your agent context engineering—not a runtime tool. It covers progressive disclosure, context degradation patterns, multi-agent architectures, memory design, and evaluation for production agents. It helps your agent design its own skills, prompts, and tools with fewer tokens. Because it uses the Open Plugins format, it can be installed in Claude Code and Cursor.
People building their own skills, agents, or MCP servers who want to reduce context costs at the design stage.
It doesn't directly save tokens; it's an information and design layer.
/plugin marketplace add muratcankoylan/Agent-Skills-for-Context-Engineering /plugin install context-engineering@context-engineering-marketplace