Build different types of Claude-powered applications — chatbots, RAG systems,
Use when working with architecture-variants patterns.
agents, content pipelines, and code generation tools.
Trigger with "claude architecture", "anthropic rag", "build with claude",
"claude agent pattern", "anthropic app design".
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
Build different types of Claude-powered applications — chatbots, RAG systems,
Use when working with architecture-variants patterns.
agents, content pipelines, and code generation tools.
Trigger with "claude architecture", "anthropic rag", "build with claude",
"claude agent pattern", "anthropic app design".
Five architecture patterns for Claude-powered applications: Chatbot (stateless API wrapper), RAG (retrieval-augmented generation with vector search), Agent (tool use loop), Content Pipeline (batch processing), and Evaluation (using Claude as a judge). Each includes complete code and a comparison table.
1. Chatbot (Stateless API Wrapper)
Simplest pattern — proxy Claude with a system prompt.
Best for: Summarization, classification, extraction at scale.
5. Evaluation / Grading
Use Claude to evaluate other AI outputs or human content.
const evaluation = await client.messages.create({
model: 'claude-opus-4-20250514', // Best judgmentmax_tokens: 1024,
system: `You are an expert evaluator. Score the response 1-5 on accuracy, relevance, and completeness. Return JSON: { "accuracy": N, "relevance": N, "completeness": N, "reasoning": "..." }`,
messages: [{
role: 'user',
content: `Question: ${question}\nResponse to evaluate: ${candidateResponse}`,
}],
});
Best for: AI output quality, content moderation, automated grading.
Choosing a Pattern
Pattern
Latency
Cost
Complexity
Chatbot
Low (streaming)
Low
Simple
RAG
Medium (embed + search + generate)
Medium
Medium
Agent
High (multi-turn)
High
Complex
Pipeline
High (async batch)
Low (50% off)
Simple
Evaluation
Medium
Varies
Simple
Output
Architecture pattern selected based on requirements
Implementation code for chosen pattern
Cost and latency characteristics understood
Scaling strategy identified (streaming for chatbots, batches for pipelines)
Error Handling
Error
Cause
Solution
API Error
Check error type and status code
See clade-common-errors
Examples
See five numbered pattern sections with complete TypeScript code, and the Choosing a Pattern comparison table with latency, cost, and complexity ratings.