In the AI era, “vibe coding” — writing code intuitively with AI assistance — has become trendy and can be highly effective in certain contexts. However, the long-term role of human developers remains uncertain. If AI eventually operates without human oversight, many of these principles may lose relevance.
At the same time, there is a growing demand for Human in the Loop (HITL) review of every line of AI-generated code to ensure quality, correctness, and control, particularly given the risks of misuse or manipulation.
From a human developer’s perspective, “developer-friendly” code prioritizes qualities that make it easier to understand, maintain, test, and extend. The following principles reflect my curated selection, with emphasis on those I consider most important:
Testability and Debuggability: Code should be designed from the outset to be testable. If a function or logic is difficult to test, it likely signals a need for redesign. This often involves mocking data to enable isolated testing across different environments. Equally important is observability — building in mechanisms to visualize behavior and confirm that the system performs as expected, rather than assuming it works correctly.
Intuitiveness and Consistency: Just as with user interfaces, the hardest challenge is not creating complexity but achieving simplicity and intuitiveness. Consistent, intuitive code reduces cognitive load, freeing developers to focus on higher-level concerns. Standardized linting rules, automatic formatting, and adherence to idempotency (ensuring predictable outputs for the same inputs) all minimize surprises during code reviews, git diffs, and long-term maintenance.
Self-Documenting and Clarity: Well-chosen variable and function names should make code largely self-explanatory. When clarity is lacking, break logic into smaller units, add specifications, integration tests, or targeted documentation. Even self-documenting code benefits from supplementary documentation to bridge the gap between conceptual understanding and actual execution flow. Examples, searchable references, and tools like Storybook (for documenting component usage and behavior) significantly improve onboarding and maintenance efficiency.
Separation of Concerns (SoC): SoC means ensuring each module, function, or class has a single, well-defined responsibility. Interpretations evolve over time and across paradigms, but the core principle remains: focused concerns lead to cleaner, more maintainable code. Modern examples include CSS Modules for styling isolation and component-based architectures in frameworks like React, where atomic utility classes enhance clarity and reusability.
Open Ecosystem and Avoidance of Over-Abstraction: Favor widely adopted open standards to reduce vendor lock-in and maintenance burden. For APIs, use OpenAPI; for web-compatible backends, leverage standards like the Fetch API, Request / Response objects, and URL handling; for observability, adopt OpenTelemetry. By building on mature, community-supported tools rather than custom abstractions, teams avoid hidden long-term costs and benefit from broader ecosystem support.