A framework by Mike Cilla · Beta

The leadership manual nobody ever gave you.

Leadership OS turns assessments, your AI conversation history, and structured reflection into four practical tools — for understanding yourself, working with your team, getting more from AI, and knowing where to grow next.

90–120 minutes · Works with any AI tool · Free
Leadership OS · Analysis Engine
claude-opus-4
3 sources · 9 constructs
How you approach problems
● fairly sure of this
When you hit something complicated, your first move is to step back and look for how the pieces connect — you're after the root cause while other people are still reacting to symptoms. It shows up everywhere, not just in strategy, but in how you describe ordinary problems too. (Leadership OS calls this systems orientation.)
How you make the call
● a likely pattern
You take your time and weigh everything — a real strength when the call is genuinely hard. The watch-out: sometimes you wait past the point where waiting helps. Your own history shows the signs were often clear well before you let yourself commit.
→ The one worth your attention
Delegation appears less constrained by trust and more constrained by personal quality standards. The growth edge is not "delegate more" — it is defining what good enough looks like before the work begins, not after.
Governing question (extracted from your inputs)
"Is this actually sound — or does it just look right?"
Sample output · Fictional persona · Generated from 3 inputs
What you walk away with

Four tools you'll
keep using.

Not a score. Not a report you read once and forget. Four tools you'll come back to — to understand yourself, work better with your team, get sharper output from AI, and keep growing — each one written in plain language and grounded in your own inputs.

What it takes
90–120 minutes, once. Most of it is reflection you'd benefit from anyway.
📋
A personality assessment — free options work. We point you to one.
💬
Some existing AI chat history — and it still works if yours is thin.
🔒
The AI tool you already use. Nothing to install. Nothing leaves your account.
Leadership OS · Analysis Engine
claude-opus-4
You
Here are my three inputs — my PrinciplesYou results, the behavioral audit from my AI history, and my six reflection responses. Please generate my Leadership Profile.
OS
Your Leadership Profile · drawn from all three of your sources

Here's what I'm seeing across your three sources. I'll tell you how sure I am of each one as we go — some of this is solid, some is more of a hunch worth sitting with.

How you make the call ● fairly sure of this

You sit with complexity longer than most people can stand to — and when the situation is genuinely murky, that patience is one of your real strengths. But there's a flip side I kept noticing: once you've said a direction out loud to people, changing your mind starts to feel like a cost you don't want to pay. The signs were often clear well before you let yourself move. (Leadership OS calls this your decision style.)

How you come across ● a likely pattern

When you write, it's clear and tight — no wasted words. The thing to watch: when you're under time pressure, that same tightness can read as cold. What feels efficient to you can land as curt to the person reading it, especially when they can't hear your tone.

→ The one worth your attention

Your instinct is to solve the problem in front of you — which is exactly why your team brings you problems instead of solving them. The growth edge isn't a skill you're missing. It's that stepping back feels like doing less, when for you it's the harder, higher-value work.

Input-informed hypotheses for reflection — not a validated assessment. Synthetic illustrative example.
A sample of the Leadership Profile, generated inside the AI tool you already use.
01
Leadership Profile
Put precise language to leadership patterns you've probably felt for years but never fully articulated — where your instincts serve you, and where they quietly work against you.
The self-understanding everything else is built on.
02
Working-With-Me Guide
Give a new teammate, report, or boss a six-month head start on working effectively with you — how you communicate, how you decide, and how to give you feedback that actually lands.
Skip the months people usually spend figuring each other out.
03
AI Calibration Document
Make every AI tool you use respond like it already knows how you think — tuned to your style and standards from the first message, instead of generic from the first message every time.
Paste it into any new chat. Reuse it for years.
04
Development Roadmap
Stop guessing where to focus. Get specific growth priorities traced to your own inputs — not generic leadership advice — with concrete practices and a 90-day focus.
Private by design. Revisit each quarter to see what's shifted.
Understand it in 60 seconds

What it is.
Why it's different.

Four ideas. No jargon. Read this first.

📥
You bring three inputs
A personality assessment. A structured reflection on how you work. And a behavioral audit from your existing AI conversation history — a record of patterns in how you think, not how you describe yourself thinking.
🧠
Your AI finds the patterns
A synthesis prompt guides your AI to identify your decision tendencies, communication patterns, energy sources, friction points, and governing questions — across all three inputs simultaneously.
📘
You get three working documents
A Leadership User Manual. A Working-With-Me Guide. An AI Calibration Document. Not a report. Documents you share, update, and return to as your leadership evolves.
Traditional assessments
Based on self-report only
Produces a report you read once
Static — a snapshot at one point in time
No connection to how you actually work
Insight fades within weeks
Leadership OS
Combines self-report with behavioral signals
Produces documents you use and update
Living — gets more accurate over time
Grounded in patterns across the input you provide
Built to persist and compound
How it's produced
📋Assessments
🔍AI Corpus
🪞Reflection
Construct Analysis
Leadership Profile

Three kinds of input about how you operate, synthesized into one profile across nine leadership constructs. See how the methodology works →

What's actually doing the work

Your AI does the thinking.
The method keeps it consistent.

Leadership OS doesn't contain a scoring algorithm. It runs inside whatever capable AI tool you already use — and its job is to make that AI reason about you the same careful way every time, instead of improvising.

Your AI's ability

Reading patterns across a lot of text, holding nuance, weighing one thing against another, writing in plain language.

+
Our guided instruction

A fixed sequence of directives — what to look at, in what order, how to weigh agreement and disagreement, when to hold back, and how to label confidence.

=
A repeatable analysis

The same structured reasoning applied to your inputs — a somewhat standardized read, run by a general-purpose model rather than a clinical instrument.

The word doing real work there is somewhat. Because the analysis runs on a general AI model, two runs won't be identical the way a scored test would be — and we say so plainly. What the guided instruction buys is consistency of method: the same questions asked in the same order, the same discipline about what the inputs can and can't support, every time. How consistency and confidence are handled →

Framework Design
Want to understand the underlying model?
The model page explains what Leadership OS is evaluating, how conclusions are formed, the developmental constructs it's designed to surface, and how it connects to established research in personality science, reflective practice, and leadership development.
View the model → See sample outputs →
How your inputs become a profile
📋
Assessments
Trait tendencies & self-report
+
🔍
Corpus
Behavioral patterns in practice
+
🪞
Reflection
Meaning-making & self-awareness
⚖️
Leadership Profile
9 constructs · Confidence-labeled
Quality control

What this is.
What it isn't.

The Leadership OS is a self-knowledge methodology, not a psychological assessment. Understanding the boundaries keeps the outputs honest and useful.

Leadership OS can honestly claim
  • Synthesizing patterns across multiple data sources
  • Surfacing hypotheses about how you operate
  • Grounding insights in validated personality science as one input
  • Producing documents that reflect your own stated experience
  • Getting more accurate as input accumulates
  • Being a structured way to make implicit patterns explicit
Leadership OS does not claim
  • Validity or reliability in a psychometric sense
  • Normative comparison to any population
  • Predictive power for job performance or outcomes
  • Clinical or therapeutic conclusions
  • That AI observations are objective or unbiased
  • That the outputs are final or definitive
The honest framing
"This is a structured way to make your own patterns visible — using the best available inputs. The outputs are hypotheses, not facts. They're only as good as the input you bring. And they're most useful when you treat them as a starting point for reflection, not a finished description of who you are."
About this framework

Built from an
experiment I ran
on myself.

I didn't set out to build a leadership methodology. I was trying to make my AI tools more useful at work.

Over 2.5 years as a VP of People Operations, I accumulated hundreds of working conversations with AI — workforce analytics, HR system design, organizational design, business cases, leadership decisions. At some point I realized I'd accidentally built something: a detailed record of how I actually think.

When I transferred that context to a new AI tool, I had to make my thinking explicit in ways it had never been before. What came out of that process was the most accurate leadership profile I'd ever received — more specific than any 360, more honest than most coaching engagements, and grounded in actual behavioral signals rather than survey responses.

The Leadership OS is an attempt to make that process replicable for anyone, using whatever AI tools they already have, in about 90–120 minutes.

Background M.S. Industrial-Organizational Psychology, San Jose State University
Experience 15 years in People Operations, workforce analytics, and HR technology
Grounding Big Five / HEXACO science, I/O Psychology, Csíkszentmihályi's Flow model
The unexpected finding
"I was attempting to train an AI. The AI ended up training me — in a very practical sense. Making my thinking legible to a system that needed explicit context forced a level of self-reflection that accumulated organizational knowledge never demands."
The methodological difference
"Every existing leadership assessment asks you to describe yourself. The corpus audit asks the system that has watched you work to describe what it saw. That's a different kind of input — and a different kind of profile."
On what this is and isn't
"This is not a validated assessment. There's no normative sample, no inter-rater reliability. I want to be honest about what it isn't. But as a methodology for generating self-knowledge, it produced something I haven't gotten from any formal process."
How it's evolving

A living methodology,
not a finished product.

Leadership OS is in active development. Here's where it stands, what's changed, and the questions still open. Think of it as release notes crossed with research notes.

Current version

Version 1.0 · June 2026

Three inputs, nine constructs, four output tools, full confidence-labeling and construct traceability. End-to-end tested in clean AI sessions.

What's recent
  • Theoretical grounding made operational inside the analysis prompts — not just cited.
  • Every development priority now traces to its originating construct, input, and confidence level.
  • Plain-language framing of what you walk away with.
Open questions
  • Does the behavioral corpus add signal beyond what assessments alone reveal?
  • Which of the four tools do people actually return to?
  • How consistent is the analysis across different AI models and runs?
Stay connected

Three ways to
follow along.

No newsletter, no funnel, no signup wall. Leadership OS doesn't collect or store your assessment results, reflections, or outputs — and that won't change. These are simply open doors if you want to keep in touch.

01
Future experiments & updates
Interested in future Leadership OS experiments, methodology updates, and new tools as they're released? Send a note with your role and what you're curious about — entirely optional.
Send a note →
02
Help improve Leadership OS
Completed the process, or reviewed the methodology? Your honest reaction is the single most valuable input right now — what was useful, what was unclear, what surprised you. Not a testimonial request. A learning one.
Share feedback →
03
Future research
Leadership OS raises questions worth studying properly. No data is being collected yet and no validation is claimed — but if you'd want to hear about future research efforts or participation opportunities, the door is open.
Express interest →

Every link above opens your own email client. You decide what to send. Nothing is tracked or stored by this site.

Start here

Before the how,
the why.

If you've read this far, the next thing worth your time isn't the instructions — it's the reason this exists at all. Most leadership development is expensive, infrequent, and gated. This is an attempt at something more accessible, and that's the part worth understanding first.

This framework is in beta. Feedback is genuinely welcome — mikejcilla@gmail.com