跪拜 Guibai
← All articles
Frontend · Backend · Artificial Intelligence

Huawei CodeArts Adds Spec-Driven Development to Counter Vibe Coding's Enterprise Gaps

By DevUI团队 ·
Read original on juejin.cn ↗ Google Translate ↗ Alt translation

Enterprise teams that adopted Vibe Coding are hitting a wall: no design docs, no traceability, and no way to verify output against requirements. A spec-first agent that reads the existing codebase, writes acceptance criteria, and generates task lists before touching code turns AI programming from a solo prototyping toy into something a compliance team can sign off on.

Summary

Vibe Coding's conversational style breaks down in enterprise settings where requirements drift, architecture erodes, and nothing is verifiable. Specification-Driven Development (SDD) is the industry's answer, and Huawei Cloud's CodeArts Code Agent (v26.6.0) has shipped a four-stage SDD pipeline: requirement specification design, implementation plan creation, coding task planning, and task execution. The agent produces structured spec documents with acceptance criteria, analyzes existing codebases to identify what can be reused versus what needs extension, and generates task lists precise down to file paths and insertion points.

A demonstration on the ruoyi-vue-pro Spring Boot project walked through adding an account lockout feature. The agent expanded a vague requirement into a full specification with state machine diagrams, identified reusable components like the existing scheduled-task framework, and produced a dependency-ordered task list. After code generation, compilation checks caught errors automatically, and the resulting behavior matched the spec's acceptance criteria.

Early feedback from Juejin and GitCode evaluators calls the document trail a hard requirement for enterprise scenarios. One practical workflow suggestion: start projects in spec-driven mode to build the skeleton, switch to exploration mode for debugging, then return to spec-driven mode for disciplined iteration.

Takeaways
Vibe Coding's lack of upfront design causes architectural drift and rework in multi-person, long-lived enterprise projects.
Huawei CodeArts Code Agent's SDD mode runs four stages: requirement spec design, implementation plan creation, coding task planning, and task execution.
The agent performs a correlation analysis against the existing codebase, categorizing functionality as already implemented, needing extension, or needing addition before generating a plan.
Generated task lists include specific file paths and code insertion points, with acceptance criteria tied back to the original spec.
A document-checking mechanism validates that generated spec documents conform to the expected template structure.
Early testers recommend a hybrid workflow: spec-driven mode for initial scaffolding and iteration, exploration mode for debugging, then back to spec-driven mode.
Conclusions

SDD is not a rejection of Vibe Coding but a recognition that the two modes solve different problems: exploration needs speed, delivery needs proof. The hybrid workflow suggestion from testers makes this explicit.

Reading the existing codebase before generating a plan is the underrated capability here. Most AI coding tools treat every task as greenfield; this one identifies what's already built and what can be reused, which is the real bottleneck in legacy enterprise systems.

The acceptance criteria embedded in each task create a built-in quality gate that Vibe Coding completely lacks. Code isn't just generated; it's checked against the spec that spawned it, closing the loop that usually requires a separate QA step.

Concepts & terms
Specification-Driven Development (SDD)
A software engineering approach where a formal specification document drives design, task planning, code generation, and testing. Developers and AI agents agree on 'what to build' before coding starts, producing traceable artifacts at each stage.
Vibe Coding
An AI-assisted programming style where developers describe what they want conversationally and the AI generates code directly. Fast for prototyping but lacks the upfront design and documentation that enterprise projects require.
Requirement and Existing Function Correlation Analysis
A pre-coding step where the AI agent scans the target codebase and classifies functionality into three buckets: already implemented, needs extension, or needs addition. This prevents redundant code generation and identifies reusable components.
Source: juejin.cn ↗ Google Translate ↗ Backup ↗