Product updates and useful context
Read updates about tools and digital products, see what has changed, and find the available sources in each article. Browse the archive pages for earlier posts.
Articles and guides
Keep reading and explore a topic
Find longer articles in the blog, explore a topic through the guides, or browse earlier news.Cohere North Mini Code Gives Agent Builders a 30B Open Coding Model
Cohere released North Mini Code, a 30B-parameter sparse MoE coding model with 3B active parameters, Apache 2.0 weights, and training focused on agentic software engineering workflows.
Read updateGitHub Security Validation Now Covers Third-Party Coding Agents
GitHub now applies CodeQL, dependency advisory checks, and secret scanning to pull requests created by third-party coding agents, extending protections that were already available for Copilot cloud agent.
Read updateChatGPT Dreaming Memory Rolls Out With New Review Controls
OpenAI is rolling out a more scalable ChatGPT memory system that synthesizes context over time, expands beyond Plus and Pro, and makes memory review more important.
Read updateClaude Fable 5: What Anthropic's Public Mythos-Class Model Changes for AI Workflows
Anthropic has released Claude Fable 5, its first generally available Mythos-class model, while keeping Claude Mythos 5 limited to approved partners. The release matters for developers because it combines long-horizon coding and knowledge-work gains with new safety fallbacks, higher pricing, and mandatory 30-day retention.
Read updateOpenAI Adds Inline Moderation Scores to API Generation Requests
OpenAI now lets developers request moderation scores inside Responses API and Chat Completions generation calls, reducing the need for a separate moderation round trip while leaving policy decisions to the application.
Read updateIntuned Turns Browser Automation Into Agent-Maintained Playwright Code
Intuned's June Hacker News launch highlights a code-first browser automation platform where an AI agent builds, deploys, and repairs Playwright-style workflows instead of running every step through a live model.
Read update