EF-Builder: From Modeling Gap to Executable Experiment
EF-Builder is a structured skill that turns the fragmented modeling problem into a guided, end-to-end workflow. Given a system description and stakeholder questions, it executes the EFSUT methodology in two gated phases: a Conceptual Phase that analyzes questions, generates candidate experimental frames, ranks them by derivability, and verifies a maximal frame lattice — followed by a Developmental Phase that produces validated Java source files ready for DEVS simulation. Each experimental frame isolates a specific concern — throughput, resilience, boundary behavior, or cross-domain handoff — and couples to the system under test through typed input/output ports, so transport, tracking, and kill-chain components can be tested independently or composed without exposing internal state. Users control scope through natural-language prompts: request a single comprehensive frame, separate frames per question, or a full characterization across all question categories. The skill enforces strict phase ordering with verification gates before any code is generated or deployed, ensuring that every experimental frame is formally grounded before it becomes software. The outcome is a repeatable bridge from stakeholder intent to accreditable simulation — directly closing the integration gap that domain-specific tools leave open.
How EF-Builder Bridges Three Worlds That Were Never Designed to Connect
Transport, tracking, and kill-chain operate on different data, different math, and different clocks — and as the infographic illustrates, today’s best tools each master one world while remaining blind to the other two. EF-Builder solves this with the two pieces of mathematics shown at the bottom of the graphic: DEVS as a common event language that makes every domain behavior a plug-compatible building block with guaranteed correctness under coupling, and EFSUT as a question-first methodology that builds the smallest simulation needed to answer each stakeholder question — delivering swappable models, single-run multi-resolution execution, and full auditability from question to result.