Guided Experience — In Development

You type
one line.
It takes it from there.

A guided version that carries the whole process for you — and then stays with you. Install it once, say one phrase, and it runs the full experience like a thoughtful coach who shows their work. Come back any time to think a real situation through against what it learned about how you lead.

Request early access Or use the free Starter Kit now →
In active testing·Built as a Claude Skill·Coming soon
Leadership OS · Guided Experience
live session
Picking up where you left off
You
Continue Leadership OS. I've got a live one — I've been offered a bigger role and I can't tell if I actually want it.
↑ A new chat, your profile dropped back in. Now it's a sounding board.
Leadership OS
Let's hold it up to your profile rather than hand you generic advice. Two things your own findings raise:

The energy question. Your profile flags that you're most alive doing the hands-on craft, and quietest in pure oversight roles. This job is mostly the second thing. Worth asking honestly: are you drawn to the title, or the work?

Your known tell. You tend to say yes to growth before checking whether it fits — it's right there in your roadmap. So I'd slow this one down more than feels natural.
Leadership OS
One honest flag: whether this specific role is right is about its details, not your patterns — I can't see those. What I can do is make sure you're choosing it for the right reasons. Want to walk through the actual day-to-day of it?
It thinks with you — and says where the profile stops and guessing begins.
01

The free Starter Kit teaches you to drive. This one drives with you.

With the free Starter Kit, you gather your inputs, follow the methodology, and run the process yourself — maximum control, and you learn how it works under the hood. The Guided Starter Kit carries the machinery for you. You answer questions; it runs the method. Same destination, very different experience — and a few things the guided engine does that the free kit can't:

A visible behavioral audit
It examines your real evidence — exported AI conversations, work artifacts, writing — and shows you every observation in a table with the supporting evidence, an alternative interpretation, and a confidence label. Nothing counts until you confirm it, row by row.
A theory-grounded roadmap
Every development priority traces its reasoning: the pattern in your evidence, the named tradition it draws on, the change mechanism that tradition actually specifies, a concrete practice, and the honest condition under which the recommendation would be wrong.
It argues with itself
Before producing anything, it challenges its own findings, downgrades at least one conclusion when the evidence is thin, and prints a self-audit. If everything comes back glowing, that's a failure — and it knows it.
Fully guided, start to finish
No prompt management, no setup beyond installing one file. One question at a time, a clear path, and two brake pedals: type GOVERNANCE CHECK at any point and it stops, audits itself against its own rules, and corrects course. Ask “Are we still doing Leadership OS?” and it names where it went off-method, retracts anything it built while drifting, and returns to the right step.
02

Where this actually is right now.

Leadership OS is in discovery, not distribution. The free Starter Kit is fully public today. The Guided Experience below is built and running as a Claude Skill, still being refined through early testing before it's released — so it isn't downloadable yet. Here's the honest current status:

How it's built

Leadership OS installs as a single Claude Skill: one root file carrying the method, plus a reference library the model loads as it needs it — nine constructs, the research foundations, the corpus protocol, the output templates, and a silent internal quality review that runs before you're shown any conclusion. Twenty-five files in total, cross-referencing each other, so the method holds its shape across a two-hour session instead of dissolving into a friendly chat.

That architecture is why it stays on-method. It's also why the whole package installs as one upload — the root file alone isn't Leadership OS.

✓ Free Starter Kit — public and self-guided
✓ Methodology — fully documented and open to critique
✓ Guided Experience — built and running as a Claude Skill (v4.0.4)
◐ Cross-platform execution — hardened, Claude-validated; other runtimes unverified
◐ Returning later — supported as a corpus source on a follow-up run, not an automated mode
○ Early pilot cohort — forming now
○ Research & validation — ongoing
Coming soon

leadership-os-v4.zip

The complete Skill — the method, the reference library, the output templates, and the regression suite it's tested against. Installs in one upload under Settings → Capabilities → Skills.

Not yet available
Coming soon

README.md

Plain-language instructions, what to expect, and the signs something's off.

Not yet available

Want to be first in line — or help test it? Request early access

In the meantime, the free Starter Kit is available now and walks you through the same methodology yourself.

03

It doesn't end at four documents. You can bring it back.

Most leadership tools hand you a report and disappear. Your Leadership OS outputs are built to be re-entered — bring your finished Profile, Roadmap, and AI Calibration Document back into a new run and it picks up from what it already established rather than rebuilding you from scratch. That's a supported pathway, not an automated mode — so once your profile exists, you can return to it any time, right inside the AI tool you already use, and pick up where a coach would.

A check-in, when something's shifted
Come back after a role change, a hard quarter, or a stretch of real-world tests. It re-reads your profile and asks what's been confirmed, what's been complicated, and what deserves another look — without rebuilding you from scratch. If little has changed, it says so honestly rather than inventing movement.
A sounding board, for a live decision
Bring a real situation that's on your desk right now — a reorg, a difficult conversation, a tradeoff you keep circling. It holds the problem up to your specific patterns and edges, not generic leadership advice, and surfaces the blind spot your own profile predicts you'll have in exactly that moment.
In the stream of where you already work
No new app, no portal, no subscription dashboard. It runs in the same AI chat you're already in every day — which is what makes returning to it feel natural instead of like one more tool to maintain.
Honest about its own edges
It thinks with you and brings real energy to the problem — but it tells you when it crosses from applying your profile into guessing about a situation's specifics. A thinking partner you can trust precisely because it's candid about what it can and can't know.

To be plain about what this is: returning later is supported as a corpus source on a follow-up run — you bring your outputs back in and it works from them. An automated continuous mode that tracks you between runs is not built, and we'd rather say so than imply it.

04

What this is — and isn't.

Leadership OS is an exploratory, self-guided reflection methodology. It is not a personality test, not therapy, and not a validated psychometric instrument — and the guided engine will tell you that itself. Its outputs are hypotheses for you to examine, each carrying one of four confidence labels, never a score, grade, or verdict. It is built to show its evidence, argue against its own conclusions, and end every document with where it's probably wrong.

Your data stays inside the AI chat you run it in. Nothing is collected, stored, or seen by anyone else. One honest limit worth knowing before you start: Leadership OS can only examine the history it can actually reach. In Claude, work kept inside a Project is visible only to sessions in that Project — so it asks up front where your working history lives and tells you what that includes and excludes. If you work across more than one AI tool, it will write you an export prompt to run in the other one and bring back. Make that choice on purpose, because a profile built on a narrow slice will still sound confident. Redact freely — the process only needs to see how you think, never confidential specifics or anything about people who haven't consented.

New to Leadership OS? Start with the free Starter Kit to learn how the methodology works — or read the FAQs first.