Transparent and Accessible ABMs with FODD: Automatic Code from Formal ODD

Agent-based models (ABMs) face a persistent transparency challenge: open-source code rarely guarantees understanding for nonprogrammers, while readable documentation may not faithfully reflect the implementation. This documentation–code gap undermines verification, peer review, and interdisciplinary collaboration—problems that are acute for multi-agent models of artificial societies. We introduce FODD, which establishes a verified link between model documentation and executable code by extending the widely used Overview, Design concepts, and Details (ODD) protocol with formal specifications that automatically generate NetLogo code, preservingastructurefamiliartodomainexperts. Wedescribe the FODD language as implemented in JetBrains MPS, covering its structure, editor, constraint mechanisms that enforce type safety and completeness, and transformation rules for deterministic code generation. Our proof-of-concept tool separates formal, informal, and derived content, derives parts of the design documentation from behavioral and initialization specifications to reduce redundancy, and explicitly links entities, attributes, and procedures to reduce conceptual ambiguity. We tested feasibility with six diverse models and evaluated the tool along executability, expressivity, and user-friendliness. The tool generates runnable models with aligned documentation, but the behavior specification language still needs higher-level abstractions to be intuitive for domain experts. FODD has the potential to make model development more efficient by reducing manual coding stages and verification effort. At its current proof-of-concept stage, FODD primarily supports experienced modelers while laying the groundwork for broader domain-expert participation in model inspection, discussion, and validation. In this sense, it advances transparency, reproducibility, and participatory modeling in computational social science. This work is licensed under a Creative Commons Attribution International 4. 0 License. Proc. of the 25th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2026), C. Amato, L. Dennis, V. Mascardi, J. Thangarajah (eds.), May 25 – 29, 2026, Paphos, Cyprus. © 2026 International Foundation for Autonomous Agents and Multiagent Systems (www. ifaamas. org). https: //doi. org/10. 65109/XDBV4230

Authors

• Themis Dimitra Xanthopoulou Umeå University
• Andreas Prinz University of Agder
• Haakon Bøthun Lunde University of Agder
• Ivan Puga-Gonzalez NORCE
• F. LeRon Shults University of Agder & NORCE

Keywords

• Agent-Based Modeling
• Multi-Agent Systems
• Automatic Code Generation
• Model Verification
• Domain-Specific Languages
• Model Documentation
• SocialSimulation
• ExecutableSpecifications
• Transparency
• ODD
• FODD