ai-llm-patterns
Anthropic Claude integration patterns: streaming, RAG with pgvector, tool use, model selection (Haiku/Sonnet/Opus), prompt caching, and cost management for AI-powered engineering.
Discover reusable agent skills, browse implementation details, and find the right skill for your workflow.
504 skills found
Anthropic Claude integration patterns: streaming, RAG with pgvector, tool use, model selection (Haiku/Sonnet/Opus), prompt caching, and cost management for AI-powered engineering.
Autonomous multi-team codebase improvement agent with specialized modes: narrow (goal-directed), broad (hypothesis-divergent), and sweep (quality-focused).
Manage a 3-node Proxmox VE cluster (Matrix/Virgo-Core) including CEPH storage, VLAN networking, and automated VM provisioning via Python, Ansible, and Terraform.
Expert assistant for the DGame Unity framework, facilitating development, architecture, hotfix, and resource management within the TEngine-based ecosystem.
Automated global intelligence aggregator for market, geopolitical, and AI news. Features RSS feed integration, real-time alert systems for critical events, and structured report generation with intelligence inference.
Manage Neovim configurations, plugin ecosystems, and custom reproducible builds.
Interactive CLI-based issue management system for tracking, planning, and executing development tasks with full CRUD capabilities.
A local RAG semantic memory system using Qdrant and Ollama. Ideal for recalling workspace files, notes, project decisions, and user preferences with high-relevance vector search.
Orchestrate multi-agent swarms using agentic-flow for parallel task execution, dynamic topology, and intelligent coordination. Ideal for building distributed AI systems and scaling complex development workflows.
A command-line tool and Expo module for interacting with Apple HealthKit, allowing you to seed, query, and verify health data in development.
Real-time e-commerce price comparison and coupon hunting across major Chinese platforms like Taobao, JD, Pinduoduo, and more.
Aggressively prune grammatical scaffolding and filler text from inputs to optimize LLM token usage while retaining core semantic content.