Skills for managing LLM context windows, attention optimization, and token budgets for effective agent systems
9 skills available
@muratcankoylan
Methods for evaluating agent performance including LLM-as-Judge patterns, metrics design, and benchmarking
@muratcankoylan
Designing memory systems for agents including working memory, long-term storage, and retrieval strategies
@muratcankoylan
Architectural patterns for multi-agent systems including orchestration, delegation, and context sharing strategies
@muratcankoylan
Understanding and mitigating context degradation, including lost-in-the-middle phenomenon and attention curve analysis
@muratcankoylan
Foundational understanding of context window management, attention mechanics, and the 'lost-in-the-middle' phenomenon
@muratcankoylan
Belief-Desire-Intention cognitive architecture for building rational agent systems with formal mental state modeling
@muratcankoylan
Ongoing optimization of agent contexts including pruning, prioritization, and dynamic context management
@muratcankoylan
Best practices for designing agent tools including schema optimization, error handling, and output formatting
@muratcankoylan
Techniques for compressing context while preserving essential information including summarization and semantic deduplication
