paperSeptember 2026UnreviewedOpen access

A framework for efficient and secure LLM agency: a case for the GraphQL paradigm

Viktor Zhakhalov

CEUR Workshop Proceedings, Vol-4260: Proceedings of the 8th Workshop for Young Scientists in Computer Science & Software Engineering (CS&SE@SW 2025)

View resource

Abstract

LLM agents must translate natural language into concrete actions on external tools. Most systems use JSON-based function calling or, more riskily, let models emit imperative code. We propose a GraphQL-first alternative that reframes tool use as typed, declarative program synthesis against a schema. This yields three measurable advantages. First, efficiency: a token–economy analysis shows that a single GraphQL query replaces multiple RPC calls, reducing request tokens from 63 to 32 and total oper

Framework mappings

OWASP Top 10 for Agentic Applications

ASI02Tool Misuse & Exploitation

MITRE ATLAS

AML.T0053AI Agent Tool Invocation

Suggested from the entry's categories.

Cite

@inproceedings{zhakhalov2026framework,
  title = {{A framework for efficient and secure LLM agency: a case for the GraphQL paradigm}},
  author = {Viktor Zhakhalov},
  year = {2026},
  month = sep,
  booktitle = {CEUR Workshop Proceedings, Vol-4260: Proceedings of the 8th Workshop for Young Scientists in Computer Science \&amp; Software Engineering (CS\&amp;SE@SW 2025)},
  doi = {10.55056/cssesw2025/paper07},
  url = {https://www.semanticscholar.org/paper/f770e3ceaac6d17508b01151897bfcaaecc595f7}
}