rc-mcp-gen: Reducing LLM Inference Costs Through Selective Tool Injection in MCP Servers

Zenodo (CERN European Organization for Nuclear Research) 2026 Preprint Open access

Read full text
Authors: Muhammad Saad
DOI: 10.5281/zenodo.18905091
Cited by: 0

Abstract

This paper presents rc-mcp-gen, a tool that reduces LLM inference costs by 71% through selective tool injection in MCP servers. Instead of loading all 31 Rocket.Chat API tools into every request context, developers can select only the endpoints their agent needs and generate a minimal ready-to-run server. Token usage drops from 2,074 to 603 per request, cutting cost from $0.00519 to $0.00151. The tool is live at rc-mcp-gen.vercel.app.

Fields of Study & Subjects

Artificial Intelligence Applications Natural Language Processing Techniques Topic Modeling

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
7 March 2026
Language
EN
Document type
Preprint