Tag
The article discusses the importance of human taste and ownership in automated software development, emphasizing the need for human oversight in design and quality checks while balancing automation with manual review.
The author audited their AI agent and found nine security flaws where guards didn't cover default paths, proposing exemption lists for build gates as a fix and asking for opinions on managing stale exemptions.
SlopScan is a Firefox extension that displays a 'slop score' letter grade for public git repositories on GitHub, Codeberg, and SourceHut, powered by a repo-slopscore server.
Vicki Boykis argues that AI-generated jargon and verbose explanations are making codebases harder for humans to navigate, and advocates for clearer, more concise writing.
The author argues that 'technical debt' is often misused by developers to describe sloppy or careless work, and suggests 'mess' is a more honest and effective term, especially when communicating with non-technical stakeholders.
The article argues that the 'AI slop' debate conflates authorship, productivity, and engineering quality, and proposes treating generative coding systems as high-throughput, error-prone producers within an engineering control loop, shifting scarce skills toward specification, verification, and accountability.
In his article "Agentic Code Quality," Addy Osmani argues that after AI agents produce code at scale, traditional manual code review can't scale, so quality assurance has to shift to harnesses, quality gates, and constraints, and he discusses autonomy and trust.
An essay encouraging developers to write code with clear intent, arguing that naming difficulties often reveal broken abstractions, and that understanding the 'why' of code matters as much as the 'what'.
Cal Newport reflects on the initial enthusiasm for AI coding tools like Claude Code and the growing disillusionment among developers due to hidden bugs, quality issues, and unsustainable workflows, arguing that outsourcing all code production to AI is not currently viable.
Linus Ekenstam discusses the gap between AI writing 75% of Google's new code and only 10% measured velocity gain, and how Sonar aims to close that gap.
Gitar is an AI-powered code review tool that automatically fixes issues it detects in code.
A large-scale empirical study analyzing 3.52 million C++ code changes in production to compare AI-generated versus human-written code quality, finding higher coupling and compute overhead but showing targeted feedback can mitigate issues.
The author explains the reverse Jevons paradox in software engineering: when the cost of making changes rises significantly due to red tape, the total volume of changes can drop to zero, leading to a fundamental halt in incremental improvements rather than a mere slowdown.
This article presents an experiment on refactoring a large codebase written by AI agents, demonstrating that refactoring reduces token consumption for future AI-driven changes. The author details the methodology and results showing a decrease in input tokens after refactoring.
Mimofan's today's update introduces several core capability improvements, including active task notification, custom agents, parallel sub-agent execution, fast mode, etc., while optimizing infrastructure and user experience.
Addy Osmani recommends using SonarSource for catching bugs and security issues on every commit, emphasizing code quality and taste.
The article discusses the challenges and best practices for writing high-quality code with AI assistance, emphasizing the need for rigorous code review and avoiding blind trust in AI-generated output.
A critique of the AI-powered software factory movement, arguing that no amount of loop engineering can overcome fundamental model-training issues, and citing evidence of declining code quality and increased incidents.
This guide explains how to evaluate the quality of AI-generated code using tests, golden datasets, reliability checks, and human review. It provides a practical workflow for catching regressions and shipping AI-assisted code with more confidence.
Prelint is a tool that prevents product drift in AI-written code, helping maintain code quality and consistency.