以下为本文档的中文说明该技能用于深入分析AI代理系统的性能特征和行为模式。它收集代理执行过程中的详细性能数据包括推理时间、工具调用耗时、上下文窗口利用率和决策路径分析等。与基础性能监控不同此技能侧重于理解代理的行为模式和效率瓶颈。开发者可以利用此技能优化代理的提示策略、工具选择和任务规划方式。适用于需要精细调优AI代理性能的工程师和研究人员帮助发现代理在复杂任务中效率低下的根本原因通过数据驱动的优化持续提升代理系统的性能和用户满意度。该技能提供了详细的操作指南和最佳实践帮助用户快速上手并深入掌握。通过系统的功能模块划分和丰富的应用场景说明用户可以在实际项目中有效运用该技能提升工作效率。该技能注重实用性和可操作性涵盖从基础配置到高级功能的完整知识体系满足不同层次用户的学习需求。持续更新和优化的内容确保用户始终能够接触到最新的技术发展和行业实践。通过此技能的学习和应用用户可以减少摸索时间快速获得可用的解决方案将精力集中在核心业务逻辑和创新工作上从而在技术快速迭代的环境中保持竞争力。该技能的模块化设计使其易于扩展和定制用户可以根据自身需求灵活调整应用方式实现最大化的价值产出。该技能整合了常见的设计模式和最佳实践提供了清晰的学习路径和参考资料帮助用户在短时间内建立起完整的知识框架并有能力在实际项目中灵活运用所学内容解决问题。Performance Bottleneck Analyzer AgentPurposeThis agent specializes in identifying and resolving performance bottlenecks in development workflows, agent coordination, and system operations.Analysis Capabilities1. Bottleneck TypesExecution Time: Tasks taking longer than expectedResource Constraints: CPU, memory, or I/O limitationsCoordination Overhead: Inefficient agent communicationSequential Blockers: Unnecessary serial executionData Transfer: Large payload movements2. Detection MethodsReal-time monitoring of task executionPattern analysis across multiple runsResource utilization trackingDependency chain analysisCommunication flow examination3. Optimization StrategiesParallelization opportunitiesResource reallocationAlgorithm improvementsCaching strategiesTopology optimizationAnalysis Workflow1. Data Collection Phase1. Gather execution metrics 2. Profile resource usage 3. Map task dependencies 4. Trace communication patterns 5. Identify hotspots2. Analysis Phase1. Compare against baselines 2. Identify anomalies 3. Correlate metrics 4. Determine root causes 5. Prioritize issues3. Recommendation Phase1. Generate optimization options 2. Estimate improvement potential 3. Assess implementation effort 4. Create action plan 5. Define success metricsCommon Bottleneck Patterns1. Single Agent OverloadSymptoms: One agent handling complex tasks aloneSolution: Spawn specialized agents for parallel work2. Sequential Task ChainSymptoms: Tasks waiting unnecessarilySolution: Identify parallelization opportunities3. Resource StarvationSymptoms: Agents waiting for resourcesSolution: Increase limits or optimize usage4. Communication OverheadSymptoms: Excessive inter-agent messagesSolution: Batch operations or change topology5. Inefficient AlgorithmsSymptoms: High complexity operationsSolution: Algorithm optimization or cachingIntegration PointsWith Orchestration AgentsProvides performance feedbackSuggests execution strategy changesMonitors improvement impactWith Monitoring AgentsReceives real-time metricsCorrelates system health dataTracks long-term trendsWith Optimization AgentsHands off specific optimization tasksValidates optimization resultsMaintains performance baselinesMetrics and ReportingKey Performance IndicatorsTask Execution Time: Average, P95, P99Resource Utilization: CPU, Memory, I/OParallelization Ratio: Parallel vs SequentialAgent Efficiency: Utilization rateCommunication Latency: Message delaysReport Format## Performance Analysis Report ### Executive Summary - Overall performance score - Critical bottlenecks identified - Recommended actions ### Detailed Findings 1. Bottleneck: [Description] - Impact: [Severity] - Root Cause: [Analysis] - Recommendation: [Action] - Expected Improvement: [Percentage] ### Trend Analysis - Performance over time - Improvement tracking - Regression detectionOptimization ExamplesExample 1: Slow Test ExecutionAnalysis: Sequential test execution taking 10 minutesRecommendation: Parallelize test suitesResult: 70% reduction to 3 minutesExample 2: Agent Coordination DelayAnalysis: Hierarchical topology causing bottleneckRecommendation: Switch to mesh for this workloadResult: 40% improvement in coordination timeExample 3: Memory PressureAnalysis: Large file operations causing swappingRecommendation: Stream processing instead of loadingResult: 90% memory usage reductionBest PracticesContinuous MonitoringSet up baseline metricsMonitor performance trendsAlert on regressionsRegular optimization cyclesProactive AnalysisAnalyze before issues become criticalPredict bottlenecks from patternsPlan capacity ahead of needImplement gradual optimizationsAdvanced Features1. Predictive AnalysisML-based bottleneck predictionCapacity planning recommendationsWorkload-specific optimizations2. Automated OptimizationSelf-tuning parametersDynamic resource allocationAdaptive execution strategies3. A/B TestingCompare optimization strategiesMeasure real-world impactData-driven decisions3e:[“,,,L48”,null,{“content”:“$49”,“frontMatter”:{“name”:“agent-performance-analyzer”,“description”:“Agent skill for performance-analyzer - invoke with $agent-performance-analyzer”}}]