Management, Orchestration and Supervision of AI-agent COmmunities for reliable AI in software engineering

Informazioni relative al progetto

MOSAICO

ID dell’accordo di sovvenzione: 101189664

DOI

10.3030/101189664

Data della firma CE 12 Novembre 2024

Data di avvio 1 Gennaio 2025

Data di completamento 31 Dicembre 2027

Finanziato da

• Digital, Industry and Space

Costo totale

Nessun dato

Contributo UE

€ 5 218 186,88

Coordinato da

INSTITUT MINES-TELECOM
France

CORDIS fornisce collegamenti ai risultati finali pubblici e alle pubblicazioni dei progetti ORIZZONTE.

I link ai risultati e alle pubblicazioni dei progetti del 7° PQ, così come i link ad alcuni tipi di risultati specifici come dataset e software, sono recuperati dinamicamente da .OpenAIRE .

Risultati finali

Documents, reports (8)

BDI Framework for Agents in AI Communities

Specification and implementation of a generic and composable AI Agent for MOSAICO, parametrized by beliefs, desires and intentions.

MOSAICO promotional material and website

Tools used for communicating the project and its results to targeted audiences (+ beta version of the website at M2).

(Initial) plan for the Dissemination, Exploitation of results & Community building

Document including the activities for the project results, D&E&C and clustering activities + community building and open source strategy.

Formalization of a Policy Language

Grammar and metamodel definition for the policy language.

Orchestrator Framework for AI Communities - Initial version

Orchestration framework including the standardised language to describe collaboration patterns (from T3.2) and a framework to enact them reliably (from T3.3).

A taxonomy of AI agents for SE

Comprehensive taxonomy classifying AI agents for SE, providing a structured framework for understanding and categorising their functionalities, capabilities, and applications

Common Policy Languages survey

Definition of common governance policy strategies.

Comparison of protocol design approaches

Comparison of the two main approaches for designing the protocol (LSP extension versus new protocol) based on the results obtained through their proof-of-concept prototypes, network performance and interoperability, extensibility, and maintainability.

Demonstrators, pilots, prototypes (2)

Key Performance Indicators of AI agents for SE

Define a set of KPIs specific to AI agents in SE, establishing measurable benchmarks to assess and compare their effectiveness and efficiency in various contexts.

MVP reference implementations of client and server

Demonstrator with the first versions of the reference client and server for the protocol, accompanied by user and developer documentation.

Pubblicazioni

Peer reviewed articles (7)

Binary and multi-class classification of Self-Admitted Technical Debt: How far can we go?

Autori: Francesca Arcelli Fontana, Juri Di Rocco, Davide Di Ruscio, Amleto Di Salle, Phuong T. Nguyen
Pubblicato in: Information and Software Technology, Numero 187, 2025, ISSN 0950-5849
Editore: Elsevier BV
DOI: 10.1016/J.INFSOF.2025.107862

Binary and multi-class classification of Self-Admitted Technical Debt: How far can we go?

Autori: Francesca Arcelli Fontana, Juri Di Rocco, Davide Di Ruscio, Amleto Di Salle, Phuong T. Nguyen
Pubblicato in: Information and Software Technology, Numero 187, 2025, ISSN 0950-5849
Editore: Elsevier BV
DOI: 10.1016/J.INFSOF.2025.107862

Information and Software Technology

Autori: Arcelli Fontana, Francesca; Di Rocco, Juri; Di Ruscio, Davide; Di Salle, Amleto; Nguyen, Phuong
Pubblicato in: Information and Software Technology, Numero Vol. 187, No. C, 2025, ISSN 0950-5849
Editore: Elsevier BV
DOI: 10.2139/SSRN.5209087

Detection of Technical Debt in Java Source Code

Autori: Nam Le Hai, Anh M. T. Bui, Phuong T. Nguyen, Davide Di Ruscio, Rick Kazman
Pubblicato in: ACM Transactions on Software Engineering and Methodology, 2026, ISSN 1049-331X
Editore: Association for Computing Machinery (ACM)
DOI: 10.1145/3801745

Many hands make light work: An LLM-based multi-agent system for detecting malicious PyPI packages

Autori: Muhammad Umar Zeshan, Motunrayo Ibiyo, Claudio Di Sipio, Phuong T. Nguyen, Davide Di Ruscio
Pubblicato in: Journal of Systems and Software, Numero 236, 2026, ISSN 0164-1212
Editore: Elsevier BV
DOI: 10.1016/J.JSS.2026.112792

Automated summarization of software documents: an LLM-based multi-agent approach

Autori: Duc S. H. Nguyen, Minh T. Nguyen, Phuong T. Nguyen, Juri Di Rocco, Davide Di Ruscio
Pubblicato in: Automated Software Engineering, Numero 33, 2026, ISSN 0928-8910
Editore: Springer Science and Business Media LLC
DOI: 10.1007/S10515-025-00588-4

Engineering LLM-based Multi-Agent Systems: A Taxonomy of Emerging Frameworks

Autori: Davide Di Ruscio, Phuong T. Nguyen, Claudio Di Sipio, Riccardo Rubei, Juri Di Rocco
Pubblicato in: IEEE Software, 2026, ISSN 0740-7459
Editore: Institute of Electrical and Electronics Engineers (IEEE)
DOI: 10.1109/MS.2026.3694089

Conference proceedings (11)

Generate with CodeXHug: A Dataset to Enhance Model Cards with Code Usage Patterns

Autori: Stefano Palombo, Claudio Di Sipio, Juri Di Rocco, Davide Di Ruscio
Pubblicato in: Lecture Notes in Computer Science, Software Engineering and Advanced Applications, 2025
Editore: Springer Nature Switzerland
DOI: 10.1007/978-3-032-04207-1_22

When Retriever Meets Generator: A Joint Model for Code Comment Generation

Autori: Tien P. T. Le, Anh M. T. Bui, Huy N. D. Pham, Alessio Bucaioni, Phuong T. Nguyen
Pubblicato in: 19th ACM/IEEE International Symposium on Empirical Software Engineering and Measurement (ESEM 2025), 2025
Editore: IEEE
DOI: 10.48550/ARXIV.2507.12558

Towards Diagram-Based Data Model Generation with LLMs

Autori: Léna Bamouh, Thibault Béziers La Fosse, Massimo Tisi
Pubblicato in: Lecture Notes in Computer Science, Advances in Conceptual Modeling, 2025
Editore: Springer Nature Switzerland
DOI: 10.1007/978-3-032-08620-4_10

Towards Automated Governance: A DSL for Human-Agent Collaboration in Software Projects

Autori: Adem Ait, Gwendal Jouneaux, Javier Luis Cánovas Izquierdo, Jordi Cabot
Pubblicato in: 2025 40th IEEE/ACM International Conference on Automated Software Engineering (ASE), 2026
Editore: IEEE
DOI: 10.1109/ASE63991.2025.00337

Bake Two Cakes with One Oven: RL for Defusing Popularity Bias and Cold-start in Third-Party Library Recommendations

Autori: Hoang Minh Vuong, Anh M. T. Bui, Phuong T. Nguyen, Davide Di Ruscio
Pubblicato in: Proceedings of the 29th International Conference on Evaluation and Assessment in Software Engineering, 2025
Editore: ACM
DOI: 10.1145/3756681.3757043

Teamwork makes the dream work: LLMs-Based Agents for GitHub README.MD Summarization

Autori: Duc S. H. Nguyen, Bach G. Truong, Phuong T. Nguyen, Juri Di Rocco, Davide Di Ruscio
Pubblicato in: Proceedings of the 33rd ACM International Conference on the Foundations of Software Engineering, 2025
Editore: ACM
DOI: 10.1145/3696630.3728511

Detecting Malicious Source Code in PyPI Packages with LLMs: Does RAG Come in Handy

Autori: Motunrayo Osatohanmen Ibiyo, Thinakone Louangdy, Phuong T. Nguyen, Claudio Di Sipio, Davide Di Ruscio
Pubblicato in: Proceedings of the 29th International Conference on Evaluation and Assessment in Software Engineering, 2025
Editore: ACM
DOI: 10.1145/3756681.3757042

Simplicity by Obfuscation: Evaluating LLM-Driven Code Transformation with Semantic Elasticity

Autori: Lorenzo De Tomasi, Claudio Di Sipio, Antinisca Di Marco, Phuong T. Nguyen
Pubblicato in: Proceedings of the 29th International Conference on Evaluation and Assessment in Software Engineering, 2025
Editore: ACM
DOI: 10.1145/3756681.3757041

Towards sustainability model cards

Autori: Gwendal Jouneaux, Jordi Cabot
Pubblicato in: SEAA 2025, 2025
Editore: arXiv
DOI: 10.48550/ARXIV.2507.19559

Larger Is Not Always Better: Leveraging Structured Code Diffs for Comment Inconsistency Detection

Autori: Phong Nguyen and Anh M. T. Bui and Phuong T. Nguyen
Pubblicato in: SANER 2026, 2025
Editore: arXiv.org
DOI: 10.48550/ARXIV.2512.19883

MOSAICO: Management, Orchestration and Supervision of AI-agent COmmunities for reliable AI in software engineering

Autori: Tisi, Massimo; Cabot, Jordi; Di Ruscio, Davide; Garcia-Dominguez, Antonio
Pubblicato in: Proceedings of the Research Projects Exhibition Papers Presented at the 37th International Conference on Advanced Information Systems Engineering (CAiSE 2025), 2025, ISSN 1613-0073
Editore: CEUR-WS