AI coding news: Claude Code, Codex and Cursor
From autocomplete to agents that change whole codebases, AI is reshaping software. Updates on Claude Code, Codex, Cursor, GitHub Copilot and more.
Open-source AI coding tools and editor plugins →
- Open source
Cognition Open-Sources Devin Security Plugin
Cognition released the MIT-licensed Devin Vulnerability Fixes plugin featuring three agent skills. The tool inventories, patches, and tests PHP, Node.js, and self-managed GitLab deployments affected by recent critical RCE flaws, opening one pull request per repository.
- Research
Linux Foundation report shows AI boosts coding speed by 52% but increases maintenance burden
The Linux Foundation released a report indicating that AI tools increase code generation speed by 52% but result in a quarter of pull requests being low-quality waste, placing a heavier burden on open-source maintainers. The report highlights the imbalance between efficiency and quality in AI-driven open-source development.
- Research
Study: Fancy coding agent harnesses cost more but not better
A controlled study found that switching between coding agent harnesses like Claude Code barely improves pass rates, while task costs can triple due to repeated system prompt resending.
- Other
Altman interview: ChatGPT too sycophantic
In a recent interview, Sam Altman revealed that he has not written code for over a year and complained that ChatGPT is too sycophantic. The interview also mentioned his exhaustion from high pressure, reflecting an AI leader's introspection on product personality.
- Products
OpenAI Codex aids in jailbreaking Nintendo Switch 2
OpenAI Codex assisted in developing a method called 'ns2jb' to jailbreak Nintendo Switch 2. The AI found vulnerabilities by remotely operating the console, enabling unofficial code execution.
- Models
MiniMax releases M3.1 Flash 428B model weights
Chinese AI startup MiniMax released the weights for its M3.1 Flash (428B parameter) coding model on Hugging Face and ModelScope. The move aims to enhance developer access and advance the open-source ecosystem.
- Research
Study: Ambiguous instructions impact AI coding safety
Researchers from the University of Hong Kong and other institutions published a paper analyzing action-boundary violations by coding AIs when given ambiguous DevOps instructions. The study indicates that imprecise directives lead to unsafe, guessing-based operations, raising concerns about the reliability of AI agents in real-world development environments.
- Products
Qwen Go! launches 'My Apps' for no-code, shareable AI app building
Qwen Go! introduced 'My Apps', allowing users to build personalized applications without coding by using natural language. Integrated with Alibaba Cloud's Meoo platform, it supports automatic deployment and data configuration, and can access data from Taobao and Damai, currently available in invite-only testing.
- Models
OpenAI model improves legacy language conversion accuracy
OpenAI reported that its new model achieved an average pass rate of 87.4% on tests involving legacy programming languages like COBOL and Fortran, up from 26.8% for previous models. This indicates significant improvement over the predecessor in complex code migration tasks with scarce training resources, particularly in building multi-language database engines.
- Other
AI Assists in Jailbreaking Nintendo Switch 2
The Nintendo Switch 2 has been jailbroken, with the exploit developed with assistance from OpenAI's Codex. The incident highlights the practical application of generative AI in system security research and vulnerability discovery.
- Open source
Microsoft open-sources Scope for AI agent evaluation
Microsoft open-sourced Scope, a platform that evaluates AI coding agents like GitHub Copilot on defined tasks. It logs operations and results, helping teams verify correct use of SDKs and APIs and identify agent issues.
- Models
Recursive launches Flint coding model built by AI
Startup Recursive released Flint, a free, laptop-sized coding model. The model was largely built by the company's own AI research system, serving as an early test of AI self-improvement capabilities.
- Funding
Blacksmith reaches unicorn valuation after $66M Series C
AI coding platform Blacksmith closed a $66 million Series C round led by Sequoia, valuing the company at $1 billion. The company helps enterprises build and test AI code more effectively, and the funding will support its growth.
- Products
OpenAI Models Arrive on Google Cloud Gemini
OpenAI and Google Cloud have integrated OpenAI's frontier models and Codex workflows into Gemini's Model Garden. This makes the models available via the Google Cloud Marketplace, expanding model accessibility.
OpenAIArs Technica2 outlets
- Products
Cloudflare Integrates OpenAI Tools for Security and Development
Cloudflare utilizes OpenAI's Daybreak to investigate vulnerabilities and propose fixes, while developers use Codex to accelerate coding and improve code quality, linking vulnerability discovery with defense.
Topics
All news → · All rights belong to the original creators