llm-app-patterns — Claude Code skill: features and installation
This skill outlines RAG designs and tool-using agent patterns. Consult it when planning document retrieval or choosing an agent architecture.
npx claude-code-templates@latest --skill ai-research/llm-app-patternsWhat llm-app-patterns does
The RAG material follows documents from ingestion through vector retrieval to an answer grounded in context. It contrasts chunking approaches and vector stores, then sketches semantic, hybrid, and multi-query retrieval. One example returns source metadata alongside the generated response.
For agents, the skill presents a ReAct loop: the model selects a tool, receives its result, then continues or produces a final answer. The sample caps the loop at 10 iterations. When that cap is reached, it returns a maximum-iterations message, so this stopping behavior matters if you adapt the pattern.
Install it with npx claude-code-templates@latest --skill ai-research/llm-app-patterns. These are design patterns and code examples; selecting a model and storage system for your application remains a project decision.
What it helps with
- design a RAG flow for document-based answers
- compare semantic retrieval with hybrid search
- choose a vector store for existing infrastructure
- structure an agent around repeated tool calls
- attach source metadata to generated answers
Example prompts for the agent
- Sketch a RAG flow for our project documents and compare semantic and hybrid retrieval.
- Design a ReAct agent for this task and identify the tools it would call.
- Show how to return a RAG answer with metadata for its retrieved sources.
How to install llm-app-patterns
1claude-code-templates
npx claude-code-templates@latest --skill ai-research/llm-app-patterns2Skills CLI (Vercel)
npx skills add davila7/claude-code-templates --skill llm-app-patterns -a claude-code3Manually: put the files in place
~/.claude/skills/llm-app-patterns/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/llm-app-patterns/
Manually: put the files in place
.agents/skills/llm-app-patterns/SKILL.mdFor all projects use ~/.agents/skills/
Manually: put the files in place
.cursor/skills/llm-app-patterns/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install davila7/claude-code-templates llm-app-patterns2Manually: put the files in place
.github/skills/llm-app-patterns/SKILL.mdPlatform CLI
gemini skills install https://github.com/davila7/claude-code-templates --path cli-tool/components/skills/ai-research/llm-app-patternsManually: put the files in place
.devin/skills/llm-app-patterns/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about llm-app-patterns
How do I install llm-app-patterns in Claude Code?
Run npx claude-code-templates@latest --skill ai-research/llm-app-patterns. This uses the claude-code-templates package installation method.
What does llm-app-patterns do?
It provides design patterns for RAG applications and agents that use tools. The examples cover retrieval strategies and a ReAct loop.
How many iterations can the example ReAct agent run?
The sample sets its iteration cap to 10. After reaching it, the agent returns a maximum-iterations message.
