Join my community / free newsletter — sign up here

OpenAI

By Greg Nowak, 10 July, 2026

OpenAI Is Retiring Agent Builder: Save the Workflow, Not Just Prompts

OpenAI retires Agent Builder on November 30, 2026. Here is what teams need to preserve, how to choose a migration path, and how to cut over safely.

Tags

By Greg Nowak, 6 July, 2026

Before OpenAI agents touch the CRM, map the boring boundaries

Before connecting AI agents to CRM or support workflows, define system ownership, permissions, approvals, prompt-injection tests, and audit trails.

Tags

By Greg Nowak, 9 June, 2026

OpenAI File Search: Internal Docs Need Governance Before Trust

OpenAI File Search makes retrieval easy to demo, but production depends on cleaner documents, metadata, vector-store structure, expiry rules, and cost control.

Tags

By Greg Nowak, 5 June, 2026

Before Your Website AI Bot Goes Live: Prompt-Injection Controls for Chat and Lead Capture

A practical rollout guide for website AI chat and lead-capture bots: prompt-injection controls, tool permissions, validation, logging, and human handoff.

Tags

By Greg Nowak, 1 June, 2026

ChatGPT Apps with Full MCP Access: Governance Comes First

Full MCP lets ChatGPT apps change live systems. Use this practical governance plan for permissions, tool design, OAuth, pilots, approvals, and rollback.

Tags

Review Greg on Google

Illustrated infographic summarizing: AI crawler policy now has verbs: separate search, RAG, and training

AI crawler policy now has verbs: separate search, RAG, and training

2026-08-02

AI crawler rules now need separate decisions for search, RAG, and training, backed by practical testing across robots.txt, CDNs, WAFs, and CMS controls.

Illustrated infographic summarizing: WordPress Supports Old PHP; Your Production Server Shouldn’t

WordPress Supports Old PHP; Your Production Server Shouldn’t

2026-08-01

WordPress still runs on legacy PHP, but compatibility is not a security policy. Build and test your upgrade path before PHP 8.2 support ends.

Illustrated infographic summarizing: The AI-built tool your team relies on needs an owner

The AI-built tool your team relies on needs an owner

2026-07-31

AI-built internal tools can become business-critical before anyone owns them. Here is how to secure, review, monitor, and retire them without blocking useful work.

Illustrated infographic summarizing: Your AI model has an expiry date: build the migration lane now

Your AI model has an expiry date: build the migration lane now

2026-07-30

AI models retire on a schedule. Learn how to map dependencies, test replacements, release safely and preserve a working rollback route.

Illustrated infographic summarizing: Copilot Has Repo-Level Metrics Now. What Should Teams Measure?

Copilot Has Repo-Level Metrics Now. What Should Teams Measure?

2026-07-29

GitHub’s repo-level Copilot metrics show where AI is active, but not whether it adds value. This scorecard connects usage with delivery, quality, and cost.

Illustrated infographic summarizing: Not Every AI Job Needs an Instant Answer: Batch the Backlog

Not Every AI Job Needs an Instant Answer: Batch the Backlog

2026-07-28

Move delay-tolerant AI work into dependable batch queues to cut processing costs without compromising quality, data controls, or urgent workflows.

Illustrated infographic summarizing: A stray Set-Cookie can waste your CDN: audit the cache at the edge

A stray Set-Cookie can waste your CDN: audit the cache at the edge

2026-07-27

Cloudflare Cache Response Rules can recover wasted CDN capacity, but first you need a route-level audit of public, personal and authenticated responses.

Illustrated infographic summarizing: Shorter TLS certificates expose every renewal you never automated

Shorter TLS certificates expose every renewal you never automated

2026-07-26

Shorter TLS lifetimes leave less room for manual handoffs and faulty deploy hooks. Build a renewal path that protects service availability.

Illustrated infographic summarizing: One Timeout, Two Orders: Make AI Actions Safe to Retry

One Timeout, Two Orders: Make AI Actions Safe to Retry

2026-07-25

A timed-out AI action may already have succeeded. Stable keys, durable ledgers, queues and stored results prevent a routine retry from duplicating real work.

Illustrated infographic summarizing: Your AI Visibility Dashboard Needs a Methodology, Not More Charts

Your AI Visibility Dashboard Needs a Methodology, Not More Charts

2026-07-24

A practical framework for measuring AI-search visibility with fixed prompts, repeated tests, separate metrics, retained evidence, and honest reporting.