Silver Manatee · AI Adoption Coaching Re: your CodePen 😏 Mid-Year 2026

You made a thing.
So I made one back.
This one's about your AI adoption.

Here's me working your styling preferences into Christina-facing docs. (Huzzah!) You've impressed your boss with your expanding AI skills. But I've still got a few more tricks to get you to black-belt status.

Power User → Builder

51 days

From our first session to your CodePen, April 23 to June 13. Nobody's rounding that up.

Governance Framework

6 criteria

You invented it to answer a colleague's question. Consultants charge for that.

Security Snapshot

3 answers out

One email answering my May 22 questions finishes the picture.

01 What you've actually built

This isn't “using AI.” It's building with it.

These three things were not true in April. You accomplished them in a single quarter.

The Build Crushing It

You build with Claude now.

In April you were a daily ChatGPT user. By June you had built your mid-year review as an animated dashboard and handed it to your boss, no document and no help. That's the jump from operating AI to creating with it.

The Governance On Track

You govern before you deploy.

A colleague asked whether a Claude Skill could leak outside FPF. You spotted the real question underneath (when does a personal build become an org resource?) and answered it with a review framework of your own design.

The Self-Audit Nobody Does This

You fact-check yourself.

You made a claim in session 2, went home and researched it yourself, and came back five days later with a refined position and four cited sources. I coach people to do that. Almost nobody does it.

02 The frontier

Three switches left to flip.

In session 2 you told me MFA was coming, and exactly how you felt about it. Your mid-year board now says the rollout landed. I'm taking your word for it. But I still owe you a security snapshot, and it's parked on three questions from my May 22 email.

Switch 01

MFA, the specifics.

Which authenticator did you land on? Is the second factor enforced for everyone, or just privileged accounts? And which Google Workspace edition is FPF on? That last one quietly changes your options.

Specifics open

Switch 02

SSO into Claude.

Are staff signing into Claude through Google, or with a separate password? Google sign-in means your Workspace security carries into Claude for free. If it's separate, SSO is included in your Team plan. I checked the current docs, and it's still there. Verify the FPF domain, add Google SAML, and one more password dies.

One toggle · big win

Switch 03

Claude's reach into Drive & Gmail.

In Google Admin, under Manage Third-Party App Access, is Claude allow-listed for Drive, Gmail, and Calendar, or gated per app in restricted mode? That setting decides what a hijacked Claude account could actually reach. A screenshot of whatever the console says is a complete answer.

Defines the blast radius
03 Your AI arc

Power user to builder in 51 days.

You've moved through three phases since April. The slope is all yours; I just kept the notes.

April 23 · Session 1

The power user

You were a daily ChatGPT user who was already troubleshooting Salesforce at 3 AM with Claude instead of crawling Reddit for four hours. You came to session 1 with a real problem, keeping the whole org on-brand without becoming the bottleneck.

ChatGPT-nativeLyra3 AM SAML fix

May · Session 2 and after

The governor

You flipped FPF's first connector live mid-session, briefed your team on what the approval setting really does, and started sending me governance questions between our calls. AI stopped being a personal tool and became org infrastructure.

First connectorTeam briefingGovernance questions

June 13 · Now

The builder

Your mid-year review went out as a self-built, animated dashboard instead of a document. You sent it to me at 2:16 in the morning, which tracks. Two sessions in, you'd lapped the curriculum.

The CodePenLapped the curriculumNight shift
04 Where this goes next

Two moves for the second half of 2026.

For you

Build your first reusable Skill.

A Skill is a workflow you build once and anyone at FPF can trigger on cue. Two candidates are already on the table. A brand-standards Skill would solve the bottleneck problem you brought to session 1, and the policy-tracker Skill is one your colleagues have already asked about.

For FPF

Put your name on the AI policy.

FPF still runs on an unwritten AI policy. We found that gap together in session 2, and the review framework you invented in May is the strongest draft material anyone there has produced. You're the natural co-author. Along the way, keep opening up connectors deliberately, with Google Drive (where FPF actually works) as the prize.

Session 3 is up next.

You're back, I'm ready, and early July works great. We'll close out the security snapshot and build your first Skill live. Grab a Zoom slot whenever works, or I'll hit you up. Tales of travel go first on the agenda.

Welcome back stateside. — Steve