Blog
Notes and analysis on AI development.
Cursor Cloud Agents Expose the New Bottleneck in Dev Environments
Cursor introduced cloud agent development environments. The coding-agent race is moving from model quality to multi-repo workspaces, secrets, audit logs, and egress control.
AWS puts Chrome policies on a leash for browser agents
AWS AgentCore Browser moves browser-agent security from prompts into Chrome enterprise policies, custom root CAs, and auditable runtime controls.
Grok Build puts xAI into the coding-agent runtime race
xAI Grok Build Early Beta shows the coding-agent market moving from model benchmarks toward plans, approvals, extensions, parallel agents, and runtime control.
ChatGPT can now read bank accounts, and personal finance AI becomes a data boundary problem
OpenAI launched a ChatGPT Finances preview with Plaid account linking. The bigger story is how AI agents earn access to sensitive financial context.
Berget Code puts a 150 euro price on sovereign AI coding
Berget Code turns the AI coding-agent market into a data-residency, jurisdiction, and procurement question for European teams.
SoundHound OASYS turns voice AI into an operations layer
SoundHound OASYS frames voice agents as a self-learning operations layer across phones, cars, kiosks, stores, and customer workflows.
GPT-5.5 crossed 50%, exposing the real bottleneck in enterprise document agents
GPT-5.5 became the first model to pass 50% on Databricks OfficeQA Pro, showing that enterprise agents still fail on parsing, retrieval, permissions, and orchestration.
Freshworks Targets the 47% of IT Tickets Filed After Hours
Freshworks AI Agent Studio turns ITSM into an execution layer for AI agents, with MCP Gateway and xLA as the control points.
Codex mobile remote access rewires the approval loop
OpenAI Codex adds mobile remote access and automation tokens, shifting coding agents toward supervised execution, workspace identity, and audit.
Honeycomb moves AI agent observability into the operations layer
Honeycomb Agent Timeline shows how AI agent operations are shifting from model quality alone to traces, cost, policy evidence, and runtime accountability.
OpenAI Codex mobile brings the approval loop to your hand
Codex is now inside the ChatGPT mobile app. AI coding agents are becoming remote work that developers supervise, steer, and approve from anywhere.
Anthropic puts a meter on Claude agents
Claude Agent SDK and claude -p usage move into separate credits on June 15, changing the economics of agent automation.