skills.rest
tranhieutt
tranhieuttCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

llm-app-patterns

Design production-ready LLM application patterns for RAG pipelines and agent workflows.

70|42|Updated Mar 27, 2026
Search Tags:#llm#rag#observability#agent-architecture#llmops#prompt-ide
One-click install
npx skills add https://github.com/tranhieutt/software_development_department --skill llm-app-patterns-tranhieutt
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: llm-app-patterns
Source: https://github.com/tranhieutt/software_development_department/tree/main/.claude/skills/llm-app-patterns
Command: npx skills add https://github.com/tranhieutt/software_development_department --skill llm-app-patterns-tranhieutt

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Too many LLM projects struggle to move from experimentation to production because they lack cohesive guidance on RAG pipelines, tool-using agent workflows, prompt IDEs, and observability practices, causing inconsistent quality and weak monitoring.

Core Features & Use Cases

  • RAG Pipeline Blueprint: Step-by-step recommendations for chunking, embedding selection, vector storage, and hybrid retrieval so responses stay grounded in your data.
  • Agent and Prompt Architecture: Covers ReAct, function calling, plan-and-execute, and multi-agent teams plus templating, chaining, and versioning practices to orchestrate complex tasks.
  • LLMOps and Production Guardrails: Provides metrics, logging, tracing, caching, rate limiting, retries, fallbacks, and evaluation frameworks to keep long-running deployments safe and observable.
  • Use Case Example: When launching a knowledge-driven LLM assistant, combine these patterns to ingest documentation, orchestrate tools, monitor latency/cost, and iterate on prompts with A/B testing.

Quick Start

Design a retrieval-augmented generation pipeline that ingests documents, compresses context, and prompts the model for grounded answers.

Frequently Asked Questions about llm-app-patterns

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I design a production-ready RAG pipeline for an enterprise LLM application?▼

To design a production-ready RAG pipeline, you need standardized patterns for chunking strategies, embedding selection, vector storage, and hybrid retrieval to ensure responses stay grounded in your data.

What's the best way to structure agent workflows for multi-tool LLM applications?▼

Structuring multi-tool LLM agent workflows requires adopting established architectures like ReAct, function calling, plan-and-execute, and multi-agent teams, combined with prompt chaining and versioning practices to orchestrate complex tasks.

How does LLMOps observability work for monitoring long-running LLM deployments?▼

LLMOps observability works by implementing production guardrails that capture evaluation metrics, logging, tracing, rate limiting, retries, and fallback models to keep long-running deployments safe and observable.

Can I use prompt IDE practices to iterate on LLM templates with A/B testing?▼

Yes, you can use prompt IDE practices to iterate on LLM templates by applying versioning, templating, and chaining techniques, which enables A/B testing to evaluate prompt performance and improve response quality.

When do I need caching and fallback models in an LLM application architecture?▼

You need caching and fallback models in LLM application architecture when deploying production guardrails to manage latency, control API costs, and ensure resilient behavior during rate limiting or unexpected model failures.

Related Skills

Explore More Software Engineering Skills→
SignmanalSignmanal
1

chroma

Store and query vector embeddings with metadata filtering for RAG applications.

Community
Intermediate
ice-droid-99ice-droid-99

querying-mlflow-metrics

Fetch aggregated MLflow trace metrics for token usage, latency, and quality scores.

Community
Intermediate
wzyxdwllwzyxdwll
14

ai

Centralize AI/LLM capability development and governance for multi-agent systems.

Community
Advanced
numman-alinumman-ali
1.0k

orchestration

Coordinate multi-agent workflows by delegating tasks to specialized agents.

Community
Advanced