Spec2COBOLRot: An Agentic-AI Degradation Loop for Realistic COBOL Corpus Generation

arXiv:2609.26835 · cs.SE, cs.AI · Submitted 2026-09-21 · Read on arXiv

cs.SE, cs.AI

Submitted: 2026-09-21

Updated: 2026-09-21

Code: https://github.com/jbespi/spec2cobolrot-samples

Terminology

• Benchmark — 5× in this paper · explained in 2 episodes
• Agentic AI — 3× in this paper · explained in 1 episode
• Generative AI — 3× in this paper · explained in 2 episodes
• Alignment — 2× in this paper · explained in 1 episode
• Average Paragraph Length — 2× in this paper
• Dead Paragraph Rate — 2× in this paper
• Fidelity — 2× in this paper · explained in 2 episodes
• Generalization — 2× in this paper · explained in 13 episodes
• Source Lines of Code — 2× in this paper
• Adaptation — 1× in this paper · explained in 1 episode
• Agentic Loop — 1× in this paper · explained in 1 episode
• Analyzer — 1× in this paper · explained in 1 episode
• Benchmarking — 1× in this paper · explained in 3 episodes
• Decoupling — 1× in this paper · explained in 2 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• Low-Resource Languages — 1× in this paper · explained in 4 episodes
• Modularity — 1× in this paper · explained in 9 episodes
• Verification — 1× in this paper · explained in 2 episodes

Sources

• Automated Testing of COBOL to Java Transformation
• X-COBOL: A Dataset of COBOL Repositories
• COBOL-Coder: Domain-Adapted Large Language Models for COBOL Code Generation and Translation
• Quality Evaluation of COBOL to Java Code Transformation
• Automated Validation of COBOL to Java Transformation
• OpenCBS: An Open-Source COBOL Defects Benchmark Suite
• Enhancing COBOL Code Explanations: A Multi-Agents Approach Using Large Language Models
• LLM-itation is the Sincerest Form of Data: Generating Synthetic Buggy Code Submissions for Computing Education
• WizardCoder: Empowering Code Large Language Models with Evol-Instruct
• Training Language Models on Synthetic Edit Sequences Improves Code Synthesis
• Impact of Comments on LLM Comprehension of Legacy Code

Related papers

• Falsification-Based Verification of LLM-Generated Optimization Models: Sound Test Batteries and Their Detection Limits
• GitSkills: A Dataset of Agent Skills on GitHub
• SABER: Benchmarking Operational Safety of LLM Coding Agents in Stateful Project Workspaces
• PackMonitor: Enabling Zero Package Hallucinations Through Decoding-Time Monitoring
• IntentCoding: Amplifying User Intent in Code Generation
• Incentives and Outcomes in Bug Bounties