paperarXivTrust 82 · PrimaryPublished 4d agoLive · 3d ago
MCP Server Architecture Patterns for LLM-Integrated Applications
The Model Context Protocol (MCP), introduced by Anthropic in November 2024, defines a standardized interface for connecting large language models (LLMs) to external tools, data sources, and services. Within months of release, hundreds of community-built MCP servers appeared on GitHub, but no software-maintenance literature has yet described how the ecosystem is being structured in production. This industry experience paper catalogues five recurring MCP server architectural patterns observed across an enumerated corpus of fifteen independently developed servers (five production servers from the
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Covers
Covers (incoming)
Implements (incoming)
Related across the graph
repotheopenco/llmgatewayrepoIBM/mcp-context-forgerepostacklok/toolhiverepoargoproj-labs/mcp-for-argocdnewsImplementing resilience patterns with Amazon Bedrock and LLM gatewaynewsRetrofit, don’t rebuild: Agentic overlays for transforming legacy enterprise servicesrepotsouth89/toolportrepohomeassistant-ai/ha-mcprepotensorlakeai/tensorlakerepoawkoy/notion-mcp-serverrepovitas/evidranewsWhat's one local AI workflow you wish you'd discovered sooner?newsX now offers an MCP server to make its platform easier for AI tools to userepoOaklight/ToolRegistrynewsA system-level approach to prompt injection: separating instruction and data channels in LLM agents [P]repobeycom/onetool-mcp