Thomas Degueule
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Thomas Degueule has a total of 58 co-authors. Here are the top co-authors by number of shared publications:
- Jean-Rémy Falleri
Univ. Bordeaux CNRS Bordeaux INP LaBRI UMR 5800 France
7 shared publications - Romain Robbes
Univ. Bordeaux CNRS Bordeaux INP LaBRI UMR 5800 France
6 shared publications - Stefano Zacchiroli
LTCI, Télécom Paris, Institut Polytechnique de Paris, Palaiseau, France
5 shared publications Benoit Combemale
4 shared publicationsOlivier Barais
4 shared publications- Benoit Combemale
University of Rennes France
3 shared publications
Latest Publications
- Top Cited PapersOpen Access
- 4 November 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Software Engineering and Methodology
- Vol. 35 (11) , 1-44
Agentic Much? Adoption of Coding Agents on GitHub
In the first half of 2025, coding agents have emerged as a category of development tools that have very quickly transitioned to the practice. Unlike “traditional” code completion LLMs such as Copilot, agents like Cursor, Claude Code, or Codex operate with high degrees of autonomy, up to generating complete pull requests starting from a developer-provided task description. This new mode of operation is poised to change the landscape in an even larger way than code completion LLMs did, making the need to study their impact critical. Also, unlike traditional LLMs, coding agents tend to leave more explicit traces in software engineering artifacts, such as co-authoring commits or pull requests. We leverage these traces to present the first large-scale study (128,018 projects) of the adoption of coding agents on GitHub, finding an estimated adoption rate of 22.20–28.66% on February 21st, 2026, which is very high for a technology only a few months old—and increasing. We carry out an in-depth study of the adopters we identified, finding that adoption is broad: it spans the entire spectrum of project maturity; it includes established organizations; and it concerns diverse programming languages or project topics. At the commit level, we find that commits assisted by coding agents are larger than commits only authored by human developers, and have a large proportion of features and bug fixes. These findings highlight the need for further investigation into the practical use of coding agents.
- 4 November 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Software Engineering and Methodology
- Vol. 35 (11) , 1-44
Top Cited PapersOpen AccessAgentic Much? Adoption of Coding Agents on GitHub
In the first half of 2025, coding agents have emerged as a category of development tools that have very quickly transitioned to the practice. Unlike “traditional” code completion LLMs such as Copilot, agents like Cursor, Claude Code, or Codex operate with high degrees of autonomy, up to generating complete pull requests starting from a developer-provided task description. This new mode of operation is poised to change the landscape in an even larger way than code completion LLMs did, making the need to study their impact critical. Also, unlike traditional LLMs, coding agents tend to leave more explicit traces in software engineering artifacts, such as co-authoring commits or pull requests. We leverage these traces to present the first large-scale study (128,018 projects) of the adoption of coding agents on GitHub, finding an estimated adoption rate of 22.20–28.66% on February 21st, 2026, which is very high for a technology only a few months old—and increasing. We carry out an in-depth study of the adopters we identified, finding that adoption is broad: it spans the entire spectrum of project maturity; it includes established organizations; and it concerns diverse programming languages or project topics. At the commit level, we find that commits assisted by coding agents are larger than commits only authored by human developers, and have a large proportion of features and bug fixes. These findings highlight the need for further investigation into the practical use of coding agents.
- Top Cited PapersOpen Access
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 496-507
Promises, Perils, and (Timely) Heuristics for Mining Coding Agent Activity
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 496-507
Top Cited PapersOpen AccessPromises, Perils, and (Timely) Heuristics for Mining Coding Agent Activity
- 2 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Composition et interopérabilité pour l'ingénierie des langages dédiés externes
Composition and interoperability for external domain-specific language engineering: Development and evolution of Domain-Specific Languages (DSLs) is becoming recurrent in the development of complex software-intensive systems. However, despite many advances in Software Language Engineering (SLE), DSLs and their tooling still suffer from substantial development costs which hamper their successful adoption in the industry. We identify two main challenges to be addressed. First, the proliferation of independently developed and constantly evolving DSLs raises the problem of interoperability between similar languages and environments. Second, since DSLs and their environments suffer from high development costs, tools and methods must be provided to assist language designers and mitigate development costs. To address these challenges, we first propose the notion of language interface. Using language interfaces, one can vary or evolve the implementation of a DSL while retaining the compatibility with the services and environments defined on its interface. Then, we present a mechanism, named model polymorphism, for manipulating models through different language interfaces. Finally, we propose a meta-language that enables language designers to reuse legacy DSLs, compose them, extend them, and customize them to meet new requirements. We implement all our contributions in a new language workbench named Melange that supports the modular definition of DSLs and the interoperability of their tooling. We evaluate the ability of Melange to solve challenging SLE scenarios.
- 2 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Composition et interopérabilité pour l'ingénierie des langages dédiés externes
Composition and interoperability for external domain-specific language engineering: Development and evolution of Domain-Specific Languages (DSLs) is becoming recurrent in the development of complex software-intensive systems. However, despite many advances in Software Language Engineering (SLE), DSLs and their tooling still suffer from substantial development costs which hamper their successful adoption in the industry. We identify two main challenges to be addressed. First, the proliferation of independently developed and constantly evolving DSLs raises the problem of interoperability between similar languages and environments. Second, since DSLs and their environments suffer from high development costs, tools and methods must be provided to assist language designers and mitigate development costs. To address these challenges, we first propose the notion of language interface. Using language interfaces, one can vary or evolve the implementation of a DSL while retaining the compatibility with the services and environments defined on its interface. Then, we present a mechanism, named model polymorphism, for manipulating models through different language interfaces. Finally, we propose a meta-language that enables language designers to reuse legacy DSLs, compose them, extend them, and customize them to meet new requirements. We implement all our contributions in a new language workbench named Melange that supports the modular definition of DSLs and the interoperability of their tooling. We evaluate the ability of Melange to solve challenging SLE scenarios.
- Open Access
DRAGON: Robust Classification for Very Large Collections of Software Repositories
The ability to automatically classify source code repositories with ''topics'' that reflect their content and purpose is very useful, especially when navigating or searching through large software collections. However, existing approaches often rely heavily on README files and other metadata, which are frequently missing, limiting their applicability in real-world large-scale settings. We present DRAGON, a repository classifier designed for very large and diverse software collections. It operates entirely on lightweight signals commonly stored in version control systems: file and directory names, and optionally the README when available. In repository classification at scale, DRAGON improves F1@5 from 54.8% to 60.8%, surpassing the state of the art. DRAGON remains effective even when README files are absent, with performance degrading by only 6% w.r.t. when they are present. This robustness makes it practical for real-world settings where documentation is sparse or inconsistent. Furthermore, many of the remaining classification errors are near misses, where predicted labels are semantically close to the correct topics. This property increases the practical value of the predictions in real-world software collections, where suggesting a few related topics can still guide search and discovery. As a byproduct of developing DRAGON, we also release the largest open dataset to date for repository classification, consisting of 825 thousand repositories with associated ground-truth topics, sourced from the Software Heritage archive, providing a foundation for future large-scale and language-agnostic research on software repository understanding.
Open AccessDRAGON: Robust Classification for Very Large Collections of Software Repositories
The ability to automatically classify source code repositories with ''topics'' that reflect their content and purpose is very useful, especially when navigating or searching through large software collections. However, existing approaches often rely heavily on README files and other metadata, which are frequently missing, limiting their applicability in real-world large-scale settings. We present DRAGON, a repository classifier designed for very large and diverse software collections. It operates entirely on lightweight signals commonly stored in version control systems: file and directory names, and optionally the README when available. In repository classification at scale, DRAGON improves F1@5 from 54.8% to 60.8%, surpassing the state of the art. DRAGON remains effective even when README files are absent, with performance degrading by only 6% w.r.t. when they are present. This robustness makes it practical for real-world settings where documentation is sparse or inconsistent. Furthermore, many of the remaining classification errors are near misses, where predicted labels are semantically close to the correct topics. This property increases the practical value of the predictions in real-world software collections, where suggesting a few related topics can still guide search and discovery. As a byproduct of developing DRAGON, we also release the largest open dataset to date for repository classification, consisting of 825 thousand repositories with associated ground-truth topics, sourced from the Software Heritage archive, providing a foundation for future large-scale and language-agnostic research on software repository understanding.
- Open Access
DRAGON: Robust Classification for Very Large Collections of Software Repositories
The ability to automatically classify source code repositories with "topics" that reflect their content and purpose is very useful, especially when navigating or searching through large software collections. However, existing approaches often rely heavily on README files and other metadata, which are frequently missing, limiting their applicability in real-world large-scale settings. We present DRAGON, a repository classifier designed for very large and diverse software collections. It operates entirely on lightweight signals commonly stored in version control systems: file and directory names, and optionally the README when available. In repository classification at scale, DRAGON improves F1@5 from 54.8% to 60.8%, surpassing the state of the art. DRAGON remains effective even when README files are absent, with performance degrading by only 6% w.r.t. when they are present. This robustness makes it practical for real-world settings where documentation is sparse or inconsistent. Furthermore, many of the remaining classification errors are near misses, where predicted labels are semantically close to the correct topics. This property increases the practical value of the predictions in real-world software collections, where suggesting a few related topics can still guide search and discovery. As a byproduct of developing DRAGON, we also release the largest open dataset to date for repository classification, consisting of 825 thousand repositories with associated ground-truth topics, sourced from the Software Heritage archive, providing a foundation for future large-scale and language-agnostic research on software repository understanding.
Open AccessDRAGON: Robust Classification for Very Large Collections of Software Repositories
The ability to automatically classify source code repositories with "topics" that reflect their content and purpose is very useful, especially when navigating or searching through large software collections. However, existing approaches often rely heavily on README files and other metadata, which are frequently missing, limiting their applicability in real-world large-scale settings. We present DRAGON, a repository classifier designed for very large and diverse software collections. It operates entirely on lightweight signals commonly stored in version control systems: file and directory names, and optionally the README when available. In repository classification at scale, DRAGON improves F1@5 from 54.8% to 60.8%, surpassing the state of the art. DRAGON remains effective even when README files are absent, with performance degrading by only 6% w.r.t. when they are present. This robustness makes it practical for real-world settings where documentation is sparse or inconsistent. Furthermore, many of the remaining classification errors are near misses, where predicted labels are semantically close to the correct topics. This property increases the practical value of the predictions in real-world software collections, where suggesting a few related topics can still guide search and discovery. As a byproduct of developing DRAGON, we also release the largest open dataset to date for repository classification, consisting of 825 thousand repositories with associated ground-truth topics, sourced from the Software Heritage archive, providing a foundation for future large-scale and language-agnostic research on software repository understanding.
- Open Access
- 22 June 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Software Engineering
- Vol. 2 (ISSTA) , 2023-2044
What Happened in This Pipeline? Diffing Build Logs with CiDiff
Continuous integration (CI) is widely used by developers to ensure the quality and reliability of their software projects. However, diagnosing a CI regression is a tedious process that involves the manual analysis of lengthy build logs. In this paper, we explore how textual differencing can support the debugging of CI regressions. As off-the-shelf diff algorithms produce suboptimal results, in this work we introduce a new diff algorithm specifically tailored to build logs called CiDiff. We evaluate CiDiff against several baselines on a novel dataset of 17 906 CI regressions, performing an accuracy study, a quantitative study and a user-study. Notably, our algorithm reduces the number of lines to inspect by about 60 % in the median case, with reasonable overhead compared to the state-of-practice LCS-diff. Finally, our algorithm is preferred by the majority of participants in 70 % of the regression cases, whereas LCS-diff is preferred in only 5 % of the cases
- 22 June 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Software Engineering
- Vol. 2 (ISSTA) , 2023-2044
Open AccessWhat Happened in This Pipeline? Diffing Build Logs with CiDiff
Continuous integration (CI) is widely used by developers to ensure the quality and reliability of their software projects. However, diagnosing a CI regression is a tedious process that involves the manual analysis of lengthy build logs. In this paper, we explore how textual differencing can support the debugging of CI regressions. As off-the-shelf diff algorithms produce suboptimal results, in this work we introduce a new diff algorithm specifically tailored to build logs called CiDiff. We evaluate CiDiff against several baselines on a novel dataset of 17 906 CI regressions, performing an accuracy study, a quantitative study and a user-study. Notably, our algorithm reduces the number of lines to inspect by about 60 % in the median case, with reasonable overhead compared to the state-of-practice LCS-diff. Finally, our algorithm is preferred by the majority of participants in 70 % of the regression cases, whereas LCS-diff is preferred in only 5 % of the cases
- Open Access
- 17 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 102-113
Automatic Classification of Software Repositories: a Systematic Mapping Study
- 17 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 102-113
Open AccessAutomatic Classification of Software Repositories: a Systematic Mapping Study
- Open Access
- 27 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 1834-1845
Balancing the Quality and Cost of Updating Dependencies
Keeping dependencies up to date is a crucial software maintenance task that requires significant effort. Developers must choose which dependencies to update, select appropriate target versions, and minimize the impact of updates in terms of breaking changes and incompatibilities. Several factors influence the choice of a new dependency version, including its freshness, popularity, absence of vulnerabilities, and compatibility. In this paper, we formulate the dependency update problem as a multi-objective optimization problem. This approach allows for updating dependencies with a global perspective, considering all direct and indirect dependencies. It also enables developers to specify their preferences regarding the quality factors to maximize and the costs to minimize when updating. The update problem is encoded as a linear program whose solution provides an optimal update strategy that aligns with developer priorities and minimizes incompatibilities. We evaluate our approach using a dataset of 107 well-tested open-source Java projects using various configurations that reflect real-world update scenarios and consider three quality metrics: dependency freshness, a time-window popularity measure, and a vulnerability score related to CVEs. Our findings indicate that our approach generates updates that compile and pass tests as well as the naive approaches typically implemented in dependency bots. Furthermore, our approach can be up to two orders of magnitude better in terms of freshness. By considering a more comprehensive concept of quality debt, which accounts for freshness, popularity, and vulnerabilities, our approach is able to reduce quality debt while maintaining reasonable memory and time consumption.
- 27 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 1834-1845
Open AccessBalancing the Quality and Cost of Updating Dependencies
Keeping dependencies up to date is a crucial software maintenance task that requires significant effort. Developers must choose which dependencies to update, select appropriate target versions, and minimize the impact of updates in terms of breaking changes and incompatibilities. Several factors influence the choice of a new dependency version, including its freshness, popularity, absence of vulnerabilities, and compatibility. In this paper, we formulate the dependency update problem as a multi-objective optimization problem. This approach allows for updating dependencies with a global perspective, considering all direct and indirect dependencies. It also enables developers to specify their preferences regarding the quality factors to maximize and the costs to minimize when updating. The update problem is encoded as a linear program whose solution provides an optimal update strategy that aligns with developer priorities and minimizes incompatibilities. We evaluate our approach using a dataset of 107 well-tested open-source Java projects using various configurations that reflect real-world update scenarios and consider three quality metrics: dependency freshness, a time-window popularity measure, and a vulnerability score related to CVEs. Our findings indicate that our approach generates updates that compile and pass tests as well as the naive approaches typically implemented in dependency bots. Furthermore, our approach can be up to two orders of magnitude better in terms of freshness. By considering a more comprehensive concept of quality debt, which accounts for freshness, popularity, and vulnerabilities, our approach is able to reduce quality debt while maintaining reasonable memory and time consumption.
- Open Access
- 15 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 426-437
Lightweight Syntactic API Usage Analysis with UCov
Designing an effective API is essential for library developers as it is the lens through which clients will judge its usability and benefits, as well as the main friction point when the library evolves. Despite its importance, defining the boundaries of an API is a challenging task, mainly due to the diverse mechanisms provided by programming languages that have non-trivial interplays. In this paper, we present a novel conceptual framework designed to assist library maintainers in understanding the interactions allowed by their APIs via the use of syntactic usage models. These customizable models enable library maintainers to improve their design ahead of release, reducing friction during evolution. The complementary syntactic usage footprints and coverage scores, inferred from client code using the API (e.g., documentation samples, tests, third-party clients), enable developers to understand in-the-wild uses of their APIs and to reflect on the adequacy of their tests and documentation. We implement these models for Java libraries in a new tool UCov and demonstrate its capabilities on three libraries exhibiting diverse styles of interaction: jsoup, commons-cli, and spark. Our exploratory case study shows that UCov provides valuable information regarding API design and fine-grained analysis of client code to identify under-tested and under-documented library code.
- 15 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 426-437
Open AccessLightweight Syntactic API Usage Analysis with UCov
Designing an effective API is essential for library developers as it is the lens through which clients will judge its usability and benefits, as well as the main friction point when the library evolves. Despite its importance, defining the boundaries of an API is a challenging task, mainly due to the diverse mechanisms provided by programming languages that have non-trivial interplays. In this paper, we present a novel conceptual framework designed to assist library maintainers in understanding the interactions allowed by their APIs via the use of syntactic usage models. These customizable models enable library maintainers to improve their design ahead of release, reducing friction during evolution. The complementary syntactic usage footprints and coverage scores, inferred from client code using the API (e.g., documentation samples, tests, third-party clients), enable developers to understand in-the-wild uses of their APIs and to reflect on the adequacy of their tests and documentation. We implement these models for Java libraries in a new tool UCov and demonstrate its capabilities on three libraries exhibiting diverse styles of interaction: jsoup, commons-cli, and spark. Our exploratory case study shows that UCov provides valuable information regarding API design and fine-grained analysis of client code to identify under-tested and under-documented library code.
- Open Access
- 15 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 246-257
What the Fix? A Study of ASATs Rule Documentation
Automatic Static Analysis Tools (ASATs) are widely used by software developers to diffuse and enforce coding practices. Yet, we know little about the documentation of ASATs, despite it being critical to learn about the coding practices in the first place. We shed light on this through several contributions. First, we analyze the documentation of more than 100 rules of 16 ASATs for multiple programming languages, and distill a taxonomy of the purposes of the documentation---What triggers a rule; Why it is important; and how to Fix an issue---and its types of contents. Then, we conduct a survey to assess the effectiveness of the documentation in terms of its goals and types of content. We highlight opportunities for improvement in ASAT documentation. In particular, we find that the Why purpose is missing in half of the rules we survey; moreover, when the Why is present, it is more likely to have quality issues than the What and the Fix.
- 15 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 246-257
Open AccessWhat the Fix? A Study of ASATs Rule Documentation
Automatic Static Analysis Tools (ASATs) are widely used by software developers to diffuse and enforce coding practices. Yet, we know little about the documentation of ASATs, despite it being critical to learn about the coding practices in the first place. We shed light on this through several contributions. First, we analyze the documentation of more than 100 rules of 16 ASATs for multiple programming languages, and distill a taxonomy of the purposes of the documentation---What triggers a rule; Why it is important; and how to Fix an issue---and its types of contents. Then, we conduct a survey to assess the effectiveness of the documentation in terms of its goals and types of content. We highlight opportunities for improvement in ASAT documentation. In particular, we find that the Why purpose is missing in half of the rules we survey; moreover, when the Why is present, it is more likely to have quality issues than the What and the Fix.