Agentic AI Patterns Reinforce Engineering Discipline

Agentic AI patterns are reinforcing core engineering discipline as modern models increase in capability, according to Paul Duvall, author of Continuous Integration: Improving Software Quality and Reducing Risk. In a recent AI DevOps Podcast, Duvall discussed how established engineering practices are being adapted through hands-on use of agentic AI in client work, emphasizing that "engineering practices are becoming even more relevant when you have AI generating code."

The shift comes as developers face unprecedented volumes of AI-generated code. Duvall observed that he is "not reviewing every line of code now" when working with AI output, as the volume of change makes this increasingly impractical. Instead, he relies on automated validation and agentic guardrails, including codified skills that allow agents to review and refine their own output.

This evolution in development workflow has led to renewed emphasis on practices like trunk-based development, committing early and often, and automated testing. Duvall explains these become essential for maintaining quality as the rate of change increases: "You're putting in mechanisms... such that the code is reviewed... but it might not be reviewed literally by you every single time."

Specification-Driven Development Takes Center Stage

A key pattern emerging in agentic AI workflows is specification-driven development. Duvall's repository includes examples of agent-readable specifications for scenarios like AWS IAM policy generation, defining expected behavior, constraints, and acceptance criteria upfront. This enables agents to generate and validate output against clear specifications.

Describing how familiar test-first patterns are being adapted to guide AI-assisted workflows, Duvall said: "I'm literally... replicating what we did with Agile and XP... it literally says red, green, refactor... I go through that process."

The emphasis on clearer specifications addresses a critical challenge in agentic development: defining intent. Duvall noted that while AI tools can generate code quickly, vague or underspecified inputs often lead to inconsistent or unpredictable results. This has led to increased focus on driving agents with structured prompts that describe intent through role, context, and constraints.

Shifting Left and Right: The Full Lifecycle Approach

Paul Stack, System Initiative's director of product, described a similar restructuring of development processes around agents. Stack's team at SWAMP, an agentic open source platform for automating and validating infrastructure, has moved to refusing pull requests in favor of GitHub Issue-based workflows that feed into specification-driven development.

"We do not accept pull requests... if you have a design... open an issue and we'll interactively walk through this and we'll design it together," Stack explained on the DevSecOps Talks podcast.

This approach extends beyond the initial development phase. Duvall emphasized the importance of shifting-right and extending feedback loops into production. He described how observability, telemetry, and even tests in production can be used to shorten feedback cycles, interpreting and sending live signals back into the development lifecycle.

Architectural Coherence and Code Standards

Maintaining architectural coherence becomes critical as agents generate code at scale. Stack emphasized the importance of providing accurate architectural patterns and practices so that agents can "produce the code in a way that was coherent with your codebase," alongside defining architecture, constraints, and testing expectations up front.

Gergely Orosz, author of The Pragmatic Engineer newsletter, discussed an open source project that refrains from merging pull requests in favor of "remixing," where contributed PRs are rebuilt by agents in line with project standards. This approach contrasts with autonomous agents using fully automated "Ralph loops," where subagents iteratively refine solutions until requirements are met.

The Future: Smaller, More Focused Teams

Looking ahead, Duvall suggested that AI may result in smaller, more focused teams, describing a move towards a "one pizza team" as coordination overhead reduces and automation increases. He positioned this as a natural evolution, similar to earlier shifts in engineering where quality is increasingly achieved through automation rather than human inspection.

Duvall's repository of agentic AI engineering patterns is continuously updated and defines structured patterns with maturity levels across development, security, and operational scenarios. The patterns include specification-driven development, codified rules and architectural constraints, atomic decomposition with parallel agents, and observable development for workflows with automated traceability.

Beyond Code: The Evolution of Engineering Identity

Acknowledging the shift beyond code-centric development, Orosz reflected that engineering identity and practice will move up a level, beyond the code itself: "I think there is something much more than coding that makes us special and I think we should cultivate that."

The convergence of these patterns suggests that agentic AI is not replacing engineering discipline but rather reinforcing it, creating systems where automated validation, clear specifications, and production feedback loops work together to maintain quality at scale. As Duvall noted, the key is putting in mechanisms that ensure quality while acknowledging that the traditional model of human review for every line of code is no longer practical in an AI-accelerated development environment.