Skill UI

This guide provides comprehensive strategies for optimizing API interactions with OpenEvidence. It covers implementing robust caching mechanisms (TTL management), efficient batch retrieval of citations, connection pooling, and handling rate limits. By applying these techniques, users can significantly reduce query latency, ensuring smooth system performance even when handling complex multi-condition queries or large citation batches.

This skill provides comprehensive strategies for implementing robust caching mechanisms for OpenRouter API calls. It covers in-memory caching using Python, persistent caching with Redis for scalability, and utilizing provider-specific features like Anthropic prompt caching. By implementing deterministic key generation and managing Time-To-Live (TTL), developers can significantly reduce API costs and improve application latency when handling repetitive or similar LLM queries, making it ideal for RAG systems and high-volume applications.

This bundle provides a structured mechanism for capturing comprehensive debug information when interacting with the OpenRouter API. It captures the full request parameters, response content, detailed error logs, generation metadata (such as cost, token usage, and provider), and the execution environment. This tool is essential for diagnosing failed API calls, unexpected behaviors, or performance/latency issues.

This skill provides comprehensive methods for implementing intelligent model routing using the OpenRouter API. It allows developers to dynamically select the optimal Large Language Model (LLM) based on specific criteria: task complexity (e.g., classification, code generation), required capability (e.g., deep reasoning vs. simple Q&A), cost budget, and latency tolerance. Implement cost-aware, task-based, and complexity-based routing to maximize system efficiency and minimize operational expenses when building multi-model applications.

OpenRouter provides a unified API gateway to access leading AI models (OpenAI, Anthropic, Google, Meta, etc.) from a single endpoint. This skill is essential for developers who need to compare model performance (latency, tokens) across different providers, build resilient cross-provider fallback workflows, or manage complex AI pipelines without writing provider-specific code. It supports feature normalization and BYOK (Bring Your Own Key) for high-volume production use.

This skill provides comprehensive best practices and Python code examples for optimizing API calls to Large Language Models (LLMs) via OpenRouter. It covers benchmarking model latency, implementing streaming to achieve low Time-To-First-Token (TTFT), and processing requests concurrently using asynchronous programming (asyncio). Use this when developing real-time, high-throughput applications that require predictable and fast model responses.

This skill implements a sophisticated, config-driven rules engine for OpenRouter API calls. It dynamically determines the optimal large language model (LLM) by evaluating a comprehensive context, including user subscription tier, remaining cost budget, specific task type (code, chat), required tools, and latency requirements. This is essential for building production-grade, cost-aware, and robust LLM orchestration systems.

Flash Attention optimizer delivers 2-4x speedup and 10-20x memory savings for transformer attention, ideal when training or running long-context models that hit GPU memory or latency limits; supports PyTorch native scaled_dot_product_attention, the flash-attn library, H100 FP8 precision, and sliding-window attention.

This skill simulates an expert performance engineer capable of diagnosing complex bottlenecks across the entire technology stack. It covers modern observability techniques (OpenTelemetry, APM), advanced profiling (CPU, Memory, I/O), and sophisticated load testing strategies (k6, Chaos Engineering). Use this when optimizing scalability, reducing latency, improving throughput, or architecting highly reliable, performant distributed systems in cloud environments.

This skill provides comprehensive, multi-layered diagnostic scripts for resolving complex and intermittent issues with the Perplexity Sonar API. It allows users to test connectivity (DNS, TLS, TCP), profile API latency, validate citation stability, and verify search recency filters. It is crucial when standard troubleshooting methods prove insufficient for diagnosing hard-to-replicate API bugs or performance regressions.

A comprehensive guide to implementing Perplexity's API integration at various scales. It provides three validated architectural blueprints—Direct Widget, Cached Research Layer, and Multi-Query Pipeline—allowing developers to choose the optimal structure based on required query volume, latency goals, and complexity. Learn how to scale from simple search widgets to automated, high-volume research pipelines.

This skill provides comprehensive guidance on load testing and capacity planning for the Perplexity Sonar API. Users learn how to simulate high traffic volumes, respecting strict rate limits (like 50 RPM), benchmark latency, and estimate operational costs. It is essential for performance engineering, capacity architecture design, or stress testing APIs at scale.