llm▌

Production-ready patterns for RAG pipelines, agent architectures, prompt management, and LLMOps monitoring. \n \n Covers five core RAG strategies: document chunking, embedding selection, retrieval methods (semantic, hybrid, multi-query, compression), and context-aware generation with citations \n Includes four agent patterns: ReAct (reasoning + acting), function calling, plan-and-execute, and multi-agent collaboration with specialized roles \n Provides prompt engineering practices: templating wi

Auto-generates a data-driven LLM usage monitoring dashboard with cost tracking, user ranking, and PM insights. \n \n Tracks token counts, API costs, and latency across OpenAI, Anthropic, Gemini, and OpenRouter using Tokuin CLI; stores metrics in JSONL format with user context and prompt categorization \n Provides two deployment options: Next.js admin dashboard with per-user drilldown pages, or lightweight single-file HTML dashboard with charts, ranking tables, and search \n Includes automatic PM

Behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity and clarity in coding practices.

Evidence-based patterns for configuring LLM parameters, based on APOLLO and Godel-Prover research.

Master comprehensive evaluation strategies for LLM applications, from automated metrics to human evaluation and A/B testing.

llm-application-dev

Access 100+ language models via inference.sh CLI.

NVIDIA's open-source library for optimizing LLM inference with state-of-the-art performance on NVIDIA GPUs.

Security rules for building secure LLM applications, based on the OWASP Top 10 for LLM Applications 2025.

Skill by ara.so — Daily 2026 Skills collection.