The LGTM (Page 86)

The LGTM

Sign in Subscribe

KPMG’s 276,000-Person Claude Rollout Is the Enterprise Agent Governance Test Case

KPMG’s 276,000-Person Claude Rollout Is the Enterprise Agent Governance Test Case

The hard part of KPMG’s Claude rollout is not the number. Yes, 276,000-plus employees is a big deployment. Yes, KPMG works across 138 countries and territories. Yes, embedding Claude inside a client-work platform built on Microsoft Azure is a serious enterprise footprint. But the interesting question is not

OpenClaw’s Codex Secrets-Audit False Positive Is a Small Bug With a Big Governance Lesson

OpenClaw’s Codex Secrets-Audit False Positive Is a Small Bug With a Big Governance Lesson

A noisy secrets scanner is not a harmless annoyance. It is a training program for ignoring alarms. Issue #84376 is small in direct severity — OpenClaw is flagging a non-secret Codex sentinel as plaintext — but the governance lesson is much larger. Security tooling only works if operators can afford to keep

OpenClaw’s Codex Runtime Placement Fix Is the Sandbox Story Coding-Agent Comparisons Usually Miss

OpenClaw’s Codex Runtime Placement Fix Is the Sandbox Story Coding-Agent Comparisons Usually Miss

The least useful way to compare coding agents is to ask which model writes better code in a benchmark. The more useful question is uglier: where does the code actually execute when the agent decides to run a shell command? PR #84377 matters because it sits exactly at that boundary.

OpenClaw’s Media-Fallback PR Shows Why Multi-Agent Delivery Needs Receipts, Not Hope

Agent platforms love to say a task is “complete.” PR #84371 is a useful reminder that complete is not a boolean. A generated image can exist on disk, the media task can have finished successfully, the requester-agent handoff can have fired, and the human can still receive nothing useful in

n8n 2.22.0 Shows AI Workflow Builders Need Evals, MCP OAuth Correctness, and Self-Healing Plans More Than Prompt Polish

n8n 2.22.0 Shows AI Workflow Builders Need Evals, MCP OAuth Correctness, and Self-Healing Plans More Than Prompt Polish

n8n 2.22.0 is what happens when “AI builds your workflow” leaves the demo booth and runs into graph validators, OAuth audiences, credential setup flows, and users who expect the thing to fix its own mistakes. The release is long, but the interesting thread is coherent: n8n is moving

Agno 2.6.8 Adds Managed-Agent Delegation, Then Quietly Fixes the Path Traversal Problems That Matter More

Agno 2.6.8 Adds Managed-Agent Delegation, Then Quietly Fixes the Path Traversal Problems That Matter More

Agno 2.6.8 has the shape of a modern agent-framework release: a shiny delegation feature on top, a pile of runtime hardening underneath, and the hardening is the part teams will be grateful for three months from now. The release adds first-party support for Google’s Antigravity API and

OpenAI Agents SDK 0.17.3 Patches the Sandbox Leaks That Make Agent Runtimes Hard to Trust

OpenAI Agents SDK 0.17.3 Patches the Sandbox Leaks That Make Agent Runtimes Hard to Trust

OpenAI’s latest Agents SDK patch is not the kind of release that wins a demo. Good. The agent stack already has enough demos. What it needs, urgently, is fewer places where credentials leak into command logs, fewer ambiguous sandbox paths, and fewer framework helpers that silently mutate the caller’