Research

Paper

AI LLM March 16, 2026

QiboAgent: a practitioner's guideline to open source assistants for Quantum Computing code development

Authors

Lorenzo Esposito, Andrea Papaluca, Stefano Carrazza

Abstract

We introduce QiboAgent, a reference implementation designed to serve as a practitioner's guideline for developing specialized coding assistants in Quantum Computing middleware. Addressing the limitations in scientific software development of general-purpose proprietary models, we explore how lightweight, open-source Large Language Models (LLMs) provided with a custom workflow architecture compare. In detail, we experiment with two complementary paradigms: a Retrieval-Augmented Generation pipeline for high-precision information retrieval, and an autonomous agentic workflow for complex software engineering tasks. We observe that this hybrid approach significantly reduces hallucination rates in code generation compared to a proprietary baseline, achieving a peak accuracy of 90.2% with relatively small open-source models of size up to 30B parameters. Furthermore, the agentic framework exhibits advanced coding capabilities, automating the resolution of maintenance issues and new features requests, or by prototyping larger-scale refactors of the codebase, such as producing a compiled Rust module with bindings of an original pure python package, Qibo in our case. The LLM workflows used for our analysis are integrated into a user interface and a Model Context Protocol server, providing an accessible tool for Qibo developers.

Metadata

arXiv ID: 2603.15538
Provider: ARXIV
Primary Category: quant-ph
Published: 2026-03-16
Fetched: 2026-03-17 06:02

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.15538v1</id>\n    <title>QiboAgent: a practitioner's guideline to open source assistants for Quantum Computing code development</title>\n    <updated>2026-03-16T17:05:06Z</updated>\n    <link href='https://arxiv.org/abs/2603.15538v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.15538v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>We introduce QiboAgent, a reference implementation designed to serve as a practitioner's guideline for developing specialized coding assistants in Quantum Computing middleware. Addressing the limitations in scientific software development of general-purpose proprietary models, we explore how lightweight, open-source Large Language Models (LLMs) provided with a custom workflow architecture compare. In detail, we experiment with two complementary paradigms: a Retrieval-Augmented Generation pipeline for high-precision information retrieval, and an autonomous agentic workflow for complex software engineering tasks. We observe that this hybrid approach significantly reduces hallucination rates in code generation compared to a proprietary baseline, achieving a peak accuracy of 90.2% with relatively small open-source models of size up to 30B parameters. Furthermore, the agentic framework exhibits advanced coding capabilities, automating the resolution of maintenance issues and new features requests, or by prototyping larger-scale refactors of the codebase, such as producing a compiled Rust module with bindings of an original pure python package, Qibo in our case. The LLM workflows used for our analysis are integrated into a user interface and a Model Context Protocol server, providing an accessible tool for Qibo developers.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='quant-ph'/>\n    <published>2026-03-16T17:05:06Z</published>\n    <arxiv:comment>13 pages, 9 figures, 1 table. Source code and deployment instructions are publicly available at https://github.com/qiboteam/qiboagent</arxiv:comment>\n    <arxiv:primary_category term='quant-ph'/>\n    <author>\n      <name>Lorenzo Esposito</name>\n    </author>\n    <author>\n      <name>Andrea Papaluca</name>\n    </author>\n    <author>\n      <name>Stefano Carrazza</name>\n    </author>\n  </entry>"
}