Devlery

Blog

Notes and analysis on AI development.

Codex now works on a locked Mac, and Goal mode redraws the agent boundary

Codex now works on a locked Mac, and Goal mode redraws the agent boundary

OpenAI added Appshots, Goal mode GA, browser annotations, locked computer use, and admin analytics to Codex. The update shows coding agents becoming longer-running workers.

Preferred sources come to AI Search, and Google gives clicks a new button

Preferred sources come to AI Search, and Google gives clicks a new button

Google is extending Preferred Sources and Highly Cited labels into AI Overviews and AI Mode. Here is what it means for AI search, publishers, and RAG product design.

React Doctor hits 11.1k stars as AI React audit layer

React Doctor hits 11.1k stars as AI React audit layer

React Doctor adds a post-generation audit loop for React code written by coding agents, scanning state, effects, performance, security, and accessibility.

The same PR, a 12.5x bill: what coding agents really cost

The same PR, a 12.5x bill: what coding agents really cost

Joule Index V0.1 adds dollars, joules, and public traces to coding-agent benchmarks, shifting the question beyond accuracy alone.

SRE agents stall at 47% on real Kubernetes incidents

SRE agents stall at 47% on real Kubernetes incidents

Artificial Analysis ITBench-AA shows that even leading SRE agents remain below 50% on Kubernetes root-cause analysis, exposing the reliability gap in operational AI.

OpenAI Adds Live Vote Counts to Election Answers

OpenAI Adds Live Vote Counts to Election Answers

OpenAI outlined its 2026 election safeguards, combining AP vote counts, voting information, Codex Security, SynthID, usage policy, and political bias evaluations.

620,000 attacks expose a 35-point safety gap in reasoning models

620,000 attacks expose a 35-point safety gap in reasoning models

TELUS Digital tested 34 AI models with more than 620,000 adversarial attacks. The benchmark shows why enterprise AI safety is now an operating discipline.

90% of PRs are agent-built, Warp exposes the new bottleneck

90% of PRs are agent-built, Warp exposes the new bottleneck

OpenAI and Warp show that the coding-agent race is shifting from code generation to open-source verification, observability, and agent orchestration.

Furiosa and Broadcom are designing a 2nm token factory

Furiosa and Broadcom are designing a 2nm token factory

FuriosaAI and Broadcom’s third-generation inference chip plan shows how agentic AI is shifting the bottleneck from raw GPU speed to token density, networking, and power.

Copilot Studio Now Clicks Apps Without APIs

Copilot Studio Now Clicks Apps Without APIs

Microsoft Copilot Studio computer use GA moves UI automation agents from demos into enterprise deployment, audit, and governance.

One API Call Now Boots Linux, the Line Google Drew with Gemini Agents

One API Call Now Boots Linux, the Line Google Drew with Gemini Agents

Google Managed Agents extends the Gemini API from model calls into sandboxed execution, shifting where agent infrastructure begins.

The 7,000-return loop behind Codex self-improving agents

The 7,000-return loop behind Codex self-improving agents

OpenAI Tax AI shows why production traces, eval sets, and practitioner feedback matter more than agent automation alone.