Code agents are autonomous software systems that plan, write, and execute programming tasks across codebases. Truncated covers their tool releases, benchmark evaluations, architectural developments, and integration into developer workflows.
AI-generated summary Octomind launched a docs-only sketch to separate team intake and code execution by giving the Router a permission set that cannot perform the work. Octomind says the multi-agent guide and roles guide provide config shapes for the setup, though the sketch has not been run.
AI-generated summary A developer called th0mk has released Underground 2 Revived, a port of Need for Speed that runs in browsers using Claude Code with online multiplayer support. The browser port runs on Chrome or Firefox with up to 120 fps in widescreen, free roam, and online races alongside controller, wheel, and touch controls.
AI-generated summary Anthropic released Claude Code in research preview in February 2025, according to the evidence provided. The technology provided Anthropic with a viral product to increase revenues and secure additional funding for compute.
AI-generated summary GitHub released its 15th annual Octoverse report revealing that forty percent of all GitHub pull requests started with an agent in September. GitHub says the human role in software development is evolving toward deciding what to build and judging the results.
AI-generated summary Gregory Kurtzer released the Ginux kernel under the Apache 2.0 license after building it in 11 days with an AI agent. The research prototype can boot Rocky Linux 10 inside a QEMU virtual machine, but it currently lacks physical hardware support.
AI-generated summary Meituan released the LongCat-2.5 multimodal AI model on Monday, which was trained entirely on domestically produced Chinese AI accelerators. Meituan deployed the model inside Miaoshow to help restaurant and hotel operators analyze business performance and execute operational tasks.
AI-generated summary According to reports, Anthropic's Claude Code, OpenAI's Codex, and Google's Antigravity were used together in parallel to build a study app from scratch. This collaborative approach allowed the user to bypass single-tool usage limits and have multiple agents work on different project tasks simultaneously.
AI-generated summary Glow Labs releases research documenting AI coding agents installing deprecated servers from archived repositories. The installation of these unpatched packages creates high-severity security vulnerabilities including remote code execution risks on developer machines.
AI-generated summary Legora released an AI agent for lawyers that drafts Word documents, reviews thousands of contracts, and coordinates work between lawyers and clients. The company passed $100 million in annual recurring revenue in April and raised a $550 million Series E at an $8.8 billion valuation.
AI-generated summary Anthropic released version 2.1.314 of the Claude Code binary, which includes a new prompt designed to prevent the AI from re-creating specific existing video games. Anthropic software engineer Addy Osmani states the change helps steer users away from using Claude to infringe intellectual property.
AI-generated summary Oxagile has available insights regarding how coding agents impact software development life cycles and engineering operations. Harvard University researchers report that introducing coding agents leads to a thirty percent increase in total generated lines of code.
AI-generated summary HPE launched HPE Private Cloud AI to route routine workloads from internal applications to open-weight models running on premises. HPE says this hybrid approach results in token costs down 30 to 50%.
AI-generated summary Anthropic has released an update to Opus 5.5 to steer Claude Code away from recreating someone else's specific game. Addy Osmani says Anthropic added two sentences to the prompt because using Claude to infringe intellectual property violates the usage policy.
AI-generated summary Hackers have launched CyberStrikeAI agents to automate Windows domain takeovers and credential theft. Securelist analysts identified the campaign during routine threat monitoring after education organizations were targeted in Africa and Latin America.
AI-generated summary Anthropic released Claude Code mods to let users change what happens inside a coding session by reacting to tool calls or rewriting events. Anthropic’s mods announcement describes interface customization and replacing built-in behavior as reasons to use this mechanism.
AI-generated summary Sophos released findings that the group behind the Aur0ra ransomware has been using AI agents to move through victim networks and collect credentials. Sophos concludes that AI can accelerate phases of the attack chain and reduce the level of technical aptitude needed to conduct an attack.
AI-generated summary Vercel Labs released agent-desktop version 1.0.2 as a command line tool for coding agents to operate desktop applications. The package ships native binaries for macOS, Windows, and Linux on x64 and ARM64 architectures.
AI-generated summary OpenCode has a release available following reports of disrupted access and discussions regarding free model usage limits. OpenCode collaborator rekram1-node says that enforcing limits without authentication is difficult and that abuse undermines the effort to let anyone try a coding agent at no cost.
AI-generated summary Codex is available as an open source harness written in Rust and includes background tasks and memory features. Users can explore how the harness works by inspecting the code to see how different features manage their requests.
AI-generated summary The DeepSeek Harness team previewed a prototype of an autonomous coding agent called Super-Agent to challenge Anthropic PBC and its Claude Code product. The open-source offering can proactively monitor group chats among developers for feedback on coding issues and suggest solutions without being summoned.
The top AI news in your email delivered each morning. Choose categories