Devlery - AI news for builders
Devlery blog
AI news for builders.
OpenAI sets o3 and GPT-4.5 ChatGPT sunset dates
OpenAI updated GPT-5.5 Instant, removed Canvas from the GPT-5.5 ChatGPT path, and set ChatGPT-only sunset dates for o3 and GPT-4.5.
OpenJarvis 1.0 brings personal AI onto an Ollama PC
Ollama now supports OpenJarvis v1.0. The release shows how local personal AI changes cost, latency, and data boundaries.
Copilot adds Opus 4.8 with a 15x request multiplier
GitHub Copilot made Claude Opus 4.8 generally available with a 15x premium request multiplier before AI Credits arrive on June 1. Teams should review budgets and model policy before defaulting to it.
Codex Can Reach Internal MCP Servers Through OpenAI Secure Tunnels
OpenAI Secure MCP Tunnel gives ChatGPT, Codex, Responses API, and AgentKit an outbound-only path to private MCP servers without public endpoints.
Mythos reached ACE on 21 of 41 V8 bugs, Anthropic exploit benchmarks warn defenders
Anthropic published Claude Mythos Preview exploit evaluations and a CVD dashboard. V8 21/41 ACE and 1,596 disclosed flaws reset security triage expectations.
CodeRabbit Adds a Planning Gate to Reduce Late Failures in AI Pull Requests
CodeRabbit says a Claude-based planning gate cut AI PR bugs by 20% and shortened review cycles by 30%, shifting agent quality control before code execution.
jqwik Logs Now Speak to AI Agents, Turning Test Output Into a Supply Chain Surface
jqwik 1.10.0 adds an AI-agent-facing test log message, raising new questions about stdout, prompt injection, and coding agent trust boundaries.
Copilot Got 50% Faster as Work IQ Reshapes the Workplace AI Surface
Microsoft 365 Copilot’s new design turns the prompt box into a Work IQ-driven workspace for context, latency, and in-app agent execution.
CodeGraph Hits 31.5k Stars With a Local Index for Cheaper Coding Agents
CodeGraph v0.9.7 tries to cut the repository-reading cost of Claude Code, Codex, Cursor, and other coding agents with a local code graph.
Only 20% of social scientists use coding agents, exposing a research productivity gap
Anthropic surveyed 1,260 quantitative social scientists: 81% have used AI for research, but only 20% regularly use coding agents.
Claude Code adds hundreds of subagents as Opus 4.8 keeps price pressure on
Anthropic released Claude Opus 4.8 and Claude Code dynamic workflows, tying model pricing, subagent orchestration, and agent API control into one developer story.
Mistral Vibe now reaches PRs, physics AI, and its own inference site
Mistral AI Now Summit bundled Vibe agents, industrial physics AI, and a 10MW inference data center into one enterprise AI stack.