The Semi-Executable Stack: Agentic Software Engineering and the Expanding Scope of SE

2026-04-16 · Source: Artificial Intelligence · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning · Depth: Advanced, quick

Summary

The paper introduces the "Semi-Executable Stack" as a six-ring diagnostic reference model to understand the expanding scope of software engineering in the age of AI-based systems. This model addresses the shift from purely executable code to semi-executable artifacts, which combine natural language, tools, workflows, control mechanisms, and organizational routines, relying on human or probabilistic interpretation rather than deterministic execution. The six rings include executable artifacts, instructional artifacts, orchestrated execution, controls, operating logic, and societal and institutional fit. The model helps identify where contributions, bottlenecks, or organizational transitions occur and their dependencies. The authors argue that AI does not diminish software engineering's relevance but rather expands its domain, reframing common objections as engineering challenges. The paper concludes with a "preserve-versus-purify" heuristic for adapting legacy software engineering processes.

Key takeaway

For AI Scientists and Research Scientists developing agentic systems, understanding the expanded definition of "software" is crucial. Your work now encompasses not just code, but also natural language instructions, workflows, and control mechanisms that rely on probabilistic interpretation. You should apply the Semi-Executable Stack model to diagnose system complexities and use the preserve-versus-purify heuristic to adapt existing engineering practices, ensuring robust and integrated AI solutions.

Key insights

Software engineering's scope expands beyond code to semi-executable artifacts, requiring new models for understanding AI's impact.

Principles

• AI expands engineering scope, not diminishes it.
• Semi-executable artifacts combine diverse elements.
• Reframing objections as engineering targets.

Method

The Semi-Executable Stack is a six-ring diagnostic model for analyzing contributions, bottlenecks, and transitions in agentic software engineering, spanning executable code to societal fit.

In practice

• Use the six-ring model for project diagnostics.
• Apply "preserve-versus-purify" to processes.

Topics

• Semi-Executable Stack
• Agentic Software Engineering
• LLM Agents
• Software Engineering Scope
• Diagnostic Reference Model

Best for: AI Scientist, Research Scientist, Software Engineer, AI Engineer, AI Architect

Related on AIssential

• AI agents aren't replacing software engineering but expanding it far beyond code, researchers argue — AI agents expand software engineering beyond code, necessitating new methods for "semi-executable artifacts."
• Beyond Correctness: Enhancing Architectural Reasoning in Code LLMs via Scalable Labeling with Agentic Judgment — An agentic LLM pipeline scalably labels architectural quality, significantly improving code LLM reasoning and patch conformance.
• The End of Software Engineering: How AI Agents Are Fundamentally Restructuring the Software Paradigm — AI agents fundamentally restructure software engineering by making code an ephemeral tool for LLM-driven reasoning, not static logic.
• Most Influential ArXiv (Software Engineering) Papers (2026-04 Version) — LLMs and AI agents are rapidly transforming software engineering, driving innovation in code generation, testing, and repair.
• AgentFactory: A Self-Evolving Framework Through Executable Subagent Accumulation and Reuse — AgentFactory uses executable Python subagent code for LLM agent self-evolution, improving reliability and efficiency.
• Article: From Prompts to Production: A Playbook for Agentic Development — Agentic AI development requires a specialized lifecycle and testing to manage nondeterministic behavior and ensure production readiness.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.