llm-app-patterns
Design LLM Application Patterns
LLM applications can fail without clear patterns for retrieval, agents, prompts, and monitoring. This skill gives practical templates for RAG, agents, prompt workflows, and LLMOps.
Install with my Agent
Copy this request to your Agent. It includes the canonical Skill page and manifest.
Review the Skillstore skill "llm-app-patterns" from https://skillstore.io/skills/sickn33-llm-app-patterns.md and its manifest at https://skillstore.io/api/skills/sickn33-llm-app-patterns/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.Your Agent should still show its plan and request any confirmation required by the security policy.
Agent-readable resources
Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.
Test it
Using "llm-app-patterns". Design a RAG chatbot for internal policies.
Expected outcome:
A RAG plan with document-aware chunking, hybrid retrieval, grounded generation, citations, and evaluation metrics.
Using "llm-app-patterns". Select an agent pattern for a support workflow.
Expected outcome:
A comparison of function calling and ReAct, with a recommendation based on tool count, validation needs, and operational risk.
Using "llm-app-patterns". Improve LLM production reliability.
Expected outcome:
A monitoring plan covering latency percentiles, error rate, hallucination checks, cost tracking, retries, caching, and fallback model rules.
Security Audit
Medium RiskMost static findings are false positives caused by Markdown code fences, reference links, generic database variables, and RAG terminology. Two semantic concerns remain in the examples: prompt/user logging without redaction and agent tool execution without explicit guardrails. No prompt injection attempt or malicious exfiltration intent was found in SKILL.md.
Confirmed security concerns (2)
Risk Factors
⚙️ External commands (37)
🌐 Network access (6)
Share & cite this report
Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.
Copy report link
https://skillstore.io/skills/sickn33-llm-app-patterns/audits/4?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown badge
[](https://skillstore.io/skills/sickn33-llm-app-patterns?utm_source=security_passport_badge)HTML badge
<a href="https://skillstore.io/skills/sickn33-llm-app-patterns?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/sickn33-llm-app-patterns/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Embed card
<iframe src="https://skillstore.io/embed/skills/sickn33-llm-app-patterns.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>Academic citations (APA · BibTeX · CFF)
APA citation
sickn33. (2026). llm-app-patterns security audit report (audit version 4) [Author version unspecified]. Skillstore. https://skillstore.io/skills/sickn33-llm-app-patterns/audits/4BibTeX citation
@techreport{sickn33-sickn33-llm-app-patterns-2026,
author = {sickn33},
title = {llm-app-patterns security audit report (audit version 4)},
institution = {Skillstore},
year = {2026},
number = {4},
url = {https://skillstore.io/skills/sickn33-llm-app-patterns/audits/4},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "llm-app-patterns security audit report (audit version 4)"
version: "unspecified"
type: report
authors:
- name: "sickn33"
date-released: "2026-07-07"
url: "https://skillstore.io/skills/sickn33-llm-app-patterns/audits/4"
identifiers:
- type: other
value: "skillstore:sickn33-llm-app-patterns:audit:4"
description: "Skillstore immutable audit report identifier"
Skillstore Score
Why this score Evidence Confidence: HighWhat You Can Build
Plan a RAG Knowledge Assistant
Choose chunking, embedding, retrieval, prompting, and citation patterns for a grounded assistant.
Select an Agent Architecture
Compare ReAct, function calling, plan-and-execute, and multi-agent designs for a task.
Define LLMOps Monitoring
Set metrics for latency, quality, cost, reliability, evaluation, retries, caching, and fallbacks.
Try These Prompts
Use the llm-app-patterns skill to recommend a RAG design for a support knowledge base. Include ingestion, retrieval, and generation choices.
Use the llm-app-patterns skill to compare ReAct, function calling, plan-and-execute, and multi-agent designs for my workflow. List tradeoffs and risks.
Use the llm-app-patterns skill to define monitoring metrics for an LLM feature. Cover quality, cost, latency, reliability, and alerting.
Use the llm-app-patterns skill to review this LLM architecture for retrieval quality, tool safety, observability, caching, rate limits, and fallback behavior.
Best Practices
- Treat all examples as starting points and add security controls before production.
- Match architecture complexity to the task, cost limits, and operational maturity.
- Measure quality, latency, cost, and safety before expanding an LLM feature.
Avoid
- Copying agent tool execution examples without allowlists or argument validation.
- Logging prompts or user identifiers without redaction, retention limits, and access controls.
- Choosing multi-agent designs when a simple RAG or function-calling flow is enough.