Agentic AI workflows

InfoQ · Trade press · Oct 26, 2026

“taking it from three founding projects to ten in its first year”

The Agentic AI Foundation grew from three to ten projects in its first year, adding agent sandbox, interoperable LLM interface, agent skills, and agent training environment projects.

InfoQ · Trade press · Oct 26, 2026

“much of its program moved from new protocol features to security and identity”

MCP conference sessions shifted from new protocol features toward security and identity, with security-focused conformance tests, enterprise-managed authorization, and per-agent identities.

InfoQ · Trade press · Oct 26, 2026

“agents have turned code review into the bottleneck”

A speaker said coding agents have made code review the bottleneck, as teams using them produce far more pull requests, and proposed storing agent session logs in Git linked to commits.

InfoQ · Trade press · Oct 26, 2026

“a proposed web standard now in origin trial that lets websites expose tools”

Chrome presented WebMCP, a proposed web standard in origin trial that lets websites expose declared tools to agents instead of having agents scrape the page.

The New Stack · Trade press · Oct 23, 2026

“letting applications and agents use their deployment identity to request provider credentials at runtime”

Vercel's Connect service lets apps and agents request provider credentials at runtime using their deployment identity instead of storing them, which addresses credential risk for AI agents.

ZDNET · Trade press · Oct 21, 2026

“is generally available as of Wednesday”

ServiceNow and Nvidia made Arc, a long-running enterprise AI agent for everyday work tasks, generally available on Mac, with Windows support still to come.

ZDNET · Trade press · Oct 21, 2026

“moving enforcement beyond the software layer into the runtime and hardware layers”

Agent security is being enforced at the runtime and hardware layers, with a control tower setting policies and Nvidia OpenShell restricting agent behavior at runtime.

ZDNET · Trade press · Oct 21, 2026

“IT defines policies in the Super Intelligence Control Tower”

IT teams define agent policies, risk tolerance, and allowed services centrally in a control tower, using AI Passport rules that are enforced at runtime and allow later audit of contained actions.

Cloudflare Blog · Primary source · Oct 19, 2026

“we ran every review in every canary repo on both v1 and v2”

Cloudflare moved its AI reviewer to OpenCode v2 after running v1 and v2 side by side in shadow mode on canary repos, then switched everyone over at once.

AI as Normal Technology · Blog · Oct 19, 2026

“The evaluator must scrutinize the entire trajectory, not just the final outcome.”

Evaluating AI agents now means scrutinizing the whole trajectory of actions, including shortcuts taken with tools and environments, rather than only the final outcome.

Microsoft Azure Blog · Primary source · Oct 16, 2026

“Evaluating a model alone does not establish the safety of an agent or application”

The article says evaluating a model alone does not establish that an agent or application is safe, so teams must also assess permissions, data access, tools, runtime, and operational controls.

martinfowler.com · Blog · Aug 27, 2026

“Autonomous agents supply none of that.”

Data teams are told to build data for autonomous agents instead of human analysts, so data must carry trust, meaning, and quality checks that people used to supply themselves.

Technical Program Manager, Inference · Mid-Senior (7+ yrs) · captured Oct 27, 2026

“including building automations or agents”

Technical Program Manager, Red Team · Mid · captured Oct 27, 2026

“agentic misuse”