A developer warns that startups shipping AI-generated code without security review risk leaking API keys, hardcoded secrets, and misconfigured environments. The post provides a seven-point checklist c…
A developer describes how combining shift-left testing with machine learning can predict where bugs will appear before code is merged. By analyzing historical defect data and commit-level features suc…
A developer lost $12,000 in fraudulent Stripe charges after an AI agent hardcoded an API key into a public GitHub commit. The incident highlights the risk of giving LLMs direct access to codebases via…
SonarSource released a SonarQube plugin for Cursor that enables in-chat code quality and security checks. The plugin connects Cursor to any SonarQube instance via the SonarQube MCP Server, allowing de…
A developer explored integrating AI code review into Bitbucket and Jira workflows, finding that tools like Qodo can add an additional validation layer beyond traditional PR summaries. The AI reviewer …
Sonar reports that AI agents generate bloated, unmaintainable code, citing a Carnegie Mellon study of 807 open-source projects using Cursor that found a 30% rise in static analysis warnings and 41% in…
A developer researching code quality tools found that methods most tools rely on, such as complexity metrics and code smells, have weaker empirical evidence than expected. The research aims to identif…
An open-source project called Guard Skills has been released as a code quality alternative designed to catch systematic failure modes in AI-generated code, such as hallucinated APIs, mock abuse, and p…
In a 2025 benchmark comparing three industry static analysis tools against three frontier LLMs on sixty-three real vulnerabilities in ten C# projects, the LLMs achieved F1 scores between 0.75 and 0.80…
Secure Code Warrior extended its AI agent at the Gartner Security & Risk Management Summit to surface real-time training insights as developers write code, linking the SCW Trust Agent to the company's…
LogiFlow CTO Kerem used SonarQube and CodeClimate to measure the technical debt created by AI-generated code, presenting data to the board that showed AI code had a cognitive complexity score of 847 v…