SWGeno 2026

Introduction

Software systems are becoming increasingly large, complex, and interconnected. Modern software is no longer a collection of isolated programs, but a living ecosystem shaped by code reuse, dependency networks, open-source communities, evolving vulnerabilities, development practices, and intelligent software agents. Understanding such systems requires more than traditional code-level analysis. It calls for new abstractions that can capture the fine-grained units, composition mechanisms, evolutionary patterns, and ecosystem relationships behind software.

The International Workshop on Software Genomics (SWGeno) aims to explore this research direction. Inspired by biological genomics, software genomics seeks to understand software through its fundamental “genes”: reusable units of functionality, behavior, dependency, vulnerability, evolution, and engineering knowledge. By identifying, representing, and analyzing these genes, software genomics provides a new perspective for studying how software is constructed, inherited, adapted, maintained, secured, and governed.

The 2nd edition of SWGeno places a special emphasis on agentic software engineering. As software agents increasingly participate in coding, testing, debugging, maintenance, security analysis, and remediation, we need to rethink what constitutes software in this new era. If software can be specified, assembled, executed, and evolved through agents, prompts, tools, workflows, models, memory, and human-agent interactions, then code may no longer be the only central carrier of software behavior. This raises a fundamental question: what is the real gene of software in the era of agentic software?

SWGeno 2026 welcomes research that investigates the genes of traditional software systems, emerging agentic software systems, and software engineering methodology itself. We are interested not only in fine-grained representations of code and dependencies, but also in the reusable principles, practices, processes, decision patterns, validation mechanisms, and feedback loops that shape how software is engineered.

The workshop will feature keynote talks, paper presentations, and interactive discussion sessions, with the goal of shaping a forward-looking research agenda for software genomics in the era of agentic software engineering.

Call for Papers

The 2nd International Workshop on Software Genomics

SWGeno 2026

Co-located with ASE 2026

The 2nd International Workshop on Software Genomics (SWGeno 2026) invites submissions on genome-inspired approaches to software analysis, management, security, evolution, and agentic software engineering.

SWGeno aims to bring together researchers and practitioners interested in understanding software systems from a fine-grained, structural, evolutionary, and ecosystem-level perspective. Inspired by biological genomics, software genomics studies software through its fundamental units, composition mechanisms, evolutionary patterns, dependency relationships, and ecosystem interactions. This perspective offers a new way to analyze complex software systems and to support the next generation of AI- and agent-driven software engineering.

In the Era of Software Agents

Software engineering is entering a new era in which autonomous and semi-autonomous agents are increasingly used to support coding, maintenance, testing, debugging, dependency management, security analysis, and remediation. However, effective software agents need more than access to raw code. They require structured and fine-grained representations of software systems, including components, dependencies, evolution histories, behaviors, and risks.

The rise of agentic software also invites us to rethink a fundamental question: if software is increasingly specified, assembled, executed, and evolved through agents, prompts, tools, workflows, models, memory, and human-agent interactions, is code still the only central carrier of software? If not, what is the real “gene” of software?

SWGeno 2026 encourages discussion on this broader question. We are interested in the genes of traditional software systems, such as code units, functions, dependencies, behaviors, vulnerabilities, and evolution patterns. We are also interested in the genes of emerging agentic software systems, such as prompts, tools, workflows, contexts, memory, APIs, models, and coordination mechanisms.

Furthermore, SWGeno welcomes work that explores the genes of software engineering methodology: the reusable principles, practices, processes, decision patterns, validation mechanisms, and feedback loops that shape how software is designed, built, tested, maintained, secured, and governed. In this sense, software genomics is not only about understanding software artifacts, but also about understanding how software engineering itself may evolve in the era of intelligent agents.

Scope and Topics

We welcome submissions on topics including, but not limited to, the following areas.

Software Genome Representations and Foundations

• Software genome representations for traditional and agentic software systems
• Fine-grained software units, structures, behaviors, and composition models
• Software genes, functional units, semantic units, and reusable building blocks
• Software genealogy, taxonomy, inheritance, mutation, and evolution modeling
• Theoretical foundations and engineering methodologies for software genomics
• Rethinking the fundamental units of software beyond source code

Agentic Software Engineering

• Agentic program understanding and navigation over large codebases
• Dependency-aware reasoning for software agents
• Agent-based software maintenance, repair, testing, debugging, and remediation
• Software genome abstractions for improving agent reliability, transparency, and controllability
• Genes of agentic software systems, including prompts, tools, workflows, memory, context, models, and coordination mechanisms
• New software development methodologies enabled by autonomous and semi-autonomous agents

Software Engineering Methodology in the Agentic Era

• Genes of software engineering methodologies, including practices, processes, decision patterns, and validation mechanisms
• Genome-inspired modeling of software development processes and engineering workflows
• Agentic reinterpretation of requirements, design, implementation, testing, debugging, review, release, and governance
• Human-agent and multi-agent collaboration patterns in software engineering
• Evolution of software engineering methods under AI- and agent-driven development

Software Security and Supply Chain Analysis

• Agentic vulnerability analysis and software supply chain security
• Vulnerability propagation and mitigation at the software genome level
• Fine-grained risk modeling across dependencies, components, and ecosystems
• Software composition analysis based on software genome representations
• Security, trustworthiness, and governance of agentic software systems
• Risk genes in code, dependencies, workflows, tools, prompts, and agent interactions

Software Evolution, Ecosystems, and Governance

• Evolutionary patterns and adaptation in software systems
• Software ecosystem analysis, governance, and health modeling
• Open-source software evolution, reuse, inheritance, and mutation
• Software genome views of technical debt, maintainability, and ecosystem risks
• Empirical studies of software evolution and ecosystem interactions
• Governance models for agentic and continuously evolving software systems

Datasets, Benchmarks, and Infrastructure

• Datasets, benchmarks, and infrastructures for software genomics
• Tools and platforms for software genome extraction, annotation, and analysis
• Benchmarks for evaluating agentic workflows over complex software systems
• Empirical studies and industrial experiences on agentic software engineering
• Evaluation methodologies for genome-inspired and agent-driven software engineering

We particularly encourage submissions that connect novel agentic workflows with principled representations of software structure and evolution. We also welcome visionary and position papers that rethink the foundations of software engineering methodology in the age of software agents.

Submission Types

SWGeno 2026 plans to accept both research papers and position papers.

• Research papers: up to 8 pages, including references
• Position papers: up to 4 pages, including references

Research papers should present original technical contributions, empirical findings, tools, datasets, or case studies. Position papers may present new visions, emerging challenges, conceptual frameworks, or early-stage ideas that can stimulate discussion at the workshop.

Submission Instructions

All submissions should follow the ASE 2026 workshop formatting and submission instructions. Submissions must present original work that is not under review elsewhere.

At least one author of each accepted paper must register for the workshop and present the work.