Three more static code analysis sensors

2026-05-27 · Source: Martin Fowler · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning · Depth: Advanced, extended

Summary

This article details experiments with various maintainability sensors for AI coding agents, focusing on static code analysis and regression testing. The author rebuilt an internal analytics dashboard using TypeScript, NextJS, and React with agents like Cursor and Claude, observing their self-correction capabilities. Key sensors included ESLint for basic linting, dependency-cruiser for enforcing modularity rules, and a custom tool for coupling metrics. An AI-driven modularity review, using Vlad Khononov's "Modularity Skills" with Claude Opus 4.7, proved effective in identifying design issues like duplicate code and inconsistent backend calls. Mutation testing with Stryker was also employed to assess test suite effectiveness, revealing gaps even with 100% coverage. The article, published with revisions on May 19, 20, and 27, 2026, highlights the potential of computational and inferential sensors to improve AI-generated code quality.

Key takeaway

For AI Engineers building with coding agents, integrating a robust sensor system is crucial for managing code maintainability. You should implement custom static analysis rules and dependency-cruiser to enforce structural quality early. Additionally, use AI-driven modularity reviews with powerful prompts to uncover deeper design issues. Crucially, deploy mutation testing to validate the effectiveness of AI-generated test suites, preventing false senses of security and reducing future technical debt.

Key insights

Computational and inferential sensors significantly enhance AI coding agent self-correction and code maintainability, especially when combined with semantic interpretation.

Principles

• Maintainability requires continuous monitoring of functional correctness, architectural fitness, and internal quality.
• Custom sensor feedback, like lint messages, guides AI agents toward self-correction.
• AI-driven semantic interpretation is crucial for complex code quality concerns like modularity.

Method

The author rebuilt an application with AI, then implemented and observed various computational (ESLint, dependency-cruiser, mutation testing) and inferential (AI modularity review, coupling analysis) sensors to provide feedback for agent self-correction.

In practice

• Customize lint messages to provide specific self-correction guidance for AI agents.
• Use dependency-cruiser to enforce layered module structures and coding conventions.
• Employ mutation testing to identify assertion gaps in AI-generated test suites.

Topics

• AI Coding Agents
• Static Code Analysis
• Code Maintainability
• Mutation Testing
• Modularity Review
• Dependency-Cruiser

Code references

• Factory-AI/eslint-plugin
• sverweij/dependency-cruiser
• vladikk/modularity

Best for: AI Engineer, Machine Learning Engineer, MLOps Engineer

Related on AIssential

• AI tools for hearing difficulties — helpful or harmful for language learning? — AI assistive technology presents a dilemma for language learners with hearing difficulties, balancing aid with skill development.
• How Enterprise Marketing Teams Are Actually Using AI Agents in 2026 — Enterprise marketing is shifting from human-assisted AI tools to autonomous AI agents that redesign entire workflows.
• [AINews] Microsoft Build: MAI-Thinking-1 and MAI Family models — Microsoft's MAI models showcase a transparent, full-stack AI strategy emphasizing clean data and local agent execution.
• Are we getting better at explaining things because of our interaction with LLMs? — Crafting clear LLM prompts may enhance human explanatory and communication abilities.
• OpenAI expands Codex with role-specific plugins to build a general-purpose app for non-developers — OpenAI is transforming Codex into a general-purpose, role-specific AI work app, expanding beyond coding for non-developers.
• How to Use Claude Managed Agents? — Anthropic's Claude Managed Agents simplifies AI agent deployment by providing a hosted, stateful, and secure execution environment.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Martin Fowler.