Full-text search across 56273 articles. Combine with topic and date filters; results sorted by relevance.
A developer warns that reviewing AI-generated code requires a separate checklist because AI code is uniformly fluent and confidently wrong, unlike human code where defects cluster around awkward sections. The developer's…
Researchers submitted ARCTIC, an AI-powered Code Critique system, to arXiv on 31 Jul 2026 (revised 14 Aug 2026), reporting that its intent prediction achieves 0.86 F1, drift detection reaches QWK = 0.907 against human an…
A new 18-chapter book, "Claude Code from Source," reverse-engineers Anthropic's Claude Code AI coding agent from the TypeScript source maps that shipped with the tool on npm, which included a sourcesContent field contain…
ToolJet, an internal tools platform, scrapped its multi-agent system built over 11 months and rebuilt around the Model Context Protocol (MCP), letting coding agents like Claude Code and Codex generate app configs instead…
A developer argues that AI coding agents have broken the long-standing link between writing code and understanding it, since agents can now ship working software that engineers never fully internalize. Citing an engineer…
Simon Willison, a software engineer and creator of Datasette, argued on the Talking Postgres podcast that lines of code can be a meaningful productivity metric for coding agents, noting that while a senior engineer might…
Meta is requiring its engineers to submit at least one AI-assisted code fix per week to improve its internal coding agent, MetaMateCR, as part of a broader plan to have AI generate 75% of code for 65% of engineers by the…
Paranoia, a new open-source tool from developer subvertnormality, provides adversarial code reviews by running one frontier coding agent against another: it integrates with Claude Code to have Codex review code, or with …
A study of 54 students found that coding agents like Cursor improve initial task completion but harm code comprehension, leaving users unprepared to extend their code without assistance. Low-effort interaction types such…
A developer has observed that junior engineers using AI coding assistants like GitHub Copilot produce code faster and with cleaner appearance, but often cannot explain or debug the code when it fails. The developer warns…
Twelve of the 13 AI coding agents reviewed on October 3, 2026 document a way to intercept a tool call before it runs, while Continue is the only one with no hook documentation, according to a survey of each vendor's hook…
Semgrep, the code security company, announced Semgrep Guardian, a security scanning tool integrated directly into AI IDEs and other AI code generation tools, to address the security risks of AI-generated code. The move c…
SkyPilot released Sandboxes, a bring-your-own-cloud code execution layer that runs untrusted agent code on a user's existing Kubernetes infrastructure instead of a hosted third-party vendor. The tool allows teams to laun…
AI agent harnesses — the tooling layer giving models file read/write and bash access — matter more to task completion than the underlying model, according to an analysis comparing Claude Code, Codex, and Hermes. The anal…
Mex 0.8.0, an open-source tool for team memory in coding agents, introduces a setup command that prepares a code graph and wiki, installs official Claude Code and Codex skills, and protects local stores from Git. The too…
An AI dark factory is a code repository that ships its own code from a spec with no human review, operating at the top of a five-level autonomy scale for coding agents. Most builders currently operate at level three, whe…
AI-generated code is creating a new form of technical debt called cognitive debt, where developers lose understanding of how and why code was written, according to a former hardware engineer. Google and Meta report high …
AI-generated code has become significantly cheaper and faster to produce, reducing the importance of raw coding skills in software development. This shift makes understanding code more expensive and dangerous, as develop…
A developer walkthrough shows how to connect Anthropic's Playwright MCP server to Claude Code so the agent reads locators from a live browser's accessibility tree instead of guessing selectors from source code. The setup…
A LogRocket article compiling five documented vibe-coding failures reports that security firm Escape scanned more than 5,600 vibe-coded apps and found over 2,000 critical vulnerabilities, while GitGuardian found AI-assis…