The Spec Growth Engine: Spec-Anchored, Code-Coupled, Drift-Enforced Architecture for AI-Assisted Software Development

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Hartwig Grabowski
• Venue: arXiv: cs.SE
• Published: 2026-06-25
• Collected: 2026-06-26
• URL: http://arxiv.org/abs/2606.27045v1
• PDF: https://arxiv.org/pdf/2606.27045v1
• DOI: UNKNOWN

Plain-Language Brief

The paper presents the Spec Growth Engine, a framework to address the structural failure modes introduced by AI coding agents in software development. These failure modes include context explosion and silent spec-code drift. The framework uses a machine-readable spec graph, a Spine context assembler, vertical-slice growth protocol, and a drift gate to solve these issues.

What the Document Says

• The paper discusses two structural failure modes in AI coding agents: context explosion and silent spec-code drift.
• The framework proposed is called the Spec Growth Engine.
• The Spec Growth Engine uses a machine-readable spec graph with nodes that carry explicit contract/design separation.
• A Spine context assembler is used to scope agent context to an ownership path.
• The vertical-slice growth protocol enforces hardest-first ordering.
• The drift gate makes spec-code divergence a blocking merge condition.

Key Concepts Mentioned

• AI coding agents
• context explosion
• silent spec-code drift
• Spec Growth Engine
• machine-readable spec graph
• Spine context assembler
• vertical-slice growth protocol
• drift gate

Claims Made by the Document

• The Spec Growth Engine addresses the failure modes introduced by AI coding agents.
• The framework synthesizes well-established software engineering principles.

Evidence Used

• The text describes how the Spec Growth Engine is designed to solve the issues of context explosion and silent spec-code drift in AI coding agents.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

AI coding agents dramatically accelerate implementation speed but introduce two structural failure modes that existing spec-driven approaches do not fully solve: (1) context explosion -- the agent must reason over an entire repository at once, degrading output quality as the context window fills; and (2) silent spec-code drift -- code evolves, the specification does not, and the divergence becomes invisible until it is costly to repair. We present the Spec Growth Engine, a lightweight framework that addresses both failure modes through a machine-readable spec graph whose nodes carry explicit contract/design separation, a Spine context assembler that scopes agent context to an ownership path, a vertical-slice growth protocol that enforces hardest-first ordering, and a drift gate that makes spec-code divergence a blocking merge condition. The design synthesises well-established software engineering principles (Parnas information hiding, C4, ADRs, Walking Skeleton, Reflexion Models, Fitness Functions) into a lean, code-coupled, machine-enforced whole -- without the overhead of heavy-weight frameworks such as RUP or MDA.