TL; DR: AI Thinking Skills for Agile Practitioners
Most agile practitioners use AI to produce outputs more quickly. Few use it to think better. This free download gives you three AI thinking skills (Socratic Explorer, Brutal Critic, Pre-Mortem) that turn Claude into a partner for diagnosing problems, stress-testing plans, and anticipating failures before they happen.
Download the AI Thinking Skills Kit
Sign up below to receive the free AI Thinking Skills Kit. The zip file includes the three Claude Skills (Socratic Explorer, Brutal Critic, Pre-mortem) and a step-by-step installation guide:
👉 Download the Socratic Explorer, Brutal Critic, Pre-Mortem Skills 👈
Turning Claude Into a Structured Thinking Partner
The AI conversation in the Agile community has split into two camps. Camp one is buying every AI certification they can find, stacking credentials like armor against obsolescence. Camp two is waiting it out, convinced that “people skills” will protect them forever.
Both camps are solving the wrong problem.
The bottleneck was never tool proficiency. The bottleneck is judgment. And judgment does not improve just because you learned to write better prompts.
I have watched this pattern repeat across every organization I have worked with in the past two years: practitioners adopt AI tools, produce outputs faster, and then discover that faster output without better thinking creates confident garbage at scale. (A fool with an LLM is still a fool, as I keep saying.)
The A3 Framework I published earlier this year addresses the decision layer: Assist, Automate, or Avoid. It helps you decide whether AI should even touch a task. But practitioners kept asking the same follow-up question: “Once I decide to use AI in ‘Assist’ mode, how do I make sure the thinking stays mine?”
Today, I am releasing the answer as a free download: three Claude Skills that turn AI into a structured thinking partner for the hardest parts of your work.
What You Get: Three Protocols for Thinking Under Pressure
These are not prompts. They are structured protocols, each one built for a specific type of thinking that agile practitioners face constantly but rarely do well: diagnosing problems, challenging plans, and anticipating failures.
Socratic Explorer: This skill walks you through any complex challenge in three phases: first, it maps the general territory (what makes this type of problem hard?), then it distills principles and decision frameworks, and finally, it applies everything to your specific situation. You control the pace. You critique and revise each phase before advancing. At the end, it generates a document capturing the entire session.
Why this matters for you: How many times have you jumped straight to a solution in a Retrospective, a stakeholder conversation, or a strategy discussion? The Socratic Explorer forces you to understand the problem before you attempt to solve it. That is the difference between an agile practitioner, who runs events, and one who changes how an organization thinks.
Brutal Critic: This skill stress-tests your plans, strategies, and reasoning. It does not validate. It does not open with praise. It identifies the gap between what you think you are saying and what you are actually proposing, finds the load-bearing assumption in your plan that could collapse, and calculates the cost of what you are avoiding. Then it tells you what someone who has already succeeded would do differently.
Why this matters for you: Every agile practitioner I know has a plan they have never pressure-tested because they are too close to it. A Sprint goal nobody challenged. A transformation roadmap nobody red-teamed, or a career strategy built on assumptions from 2019. The Brutal Critic does what your colleagues are too polite to do.
Pre-mortem: This skill assumes your plan has already failed and works backward to find the cause. It produces a forensic report: the three most likely causes of death (ranked by probability, not drama), the hidden dependency you never tested, a modification that makes the failure survivable, and a 30-day tripwire so you know early if the failure is already in motion.
Why this matters for you: Research by Mitchell, Russo, and Pennington found that imagining an event has already occurred increases the ability to identify reasons for future outcomes by 30%. [1] Gary Klein built the pre-mortem technique on this finding. The skill exploits a cognitive asymmetry most planning processes ignore. Before you commit resources to a Sprint goal, a product strategy, a reorganization, or your next career move, you run the autopsy first.
Why Judgment Is the Real AI Skill
I have argued in two recent articles (“Why Agile Practitioners Should Be Optimistic for 2026,” Part 1 and Part 2) that organizations adopting AI are failing for the same structural reasons they failed at Agile transformations. They perform change instead of actually changing. They buy tools before identifying problems. They measure adoption dashboards while business outcomes stay flat.
The organizations that succeed need people who can do the things AI cannot: read the room, sense when a stated agreement hides real resistance, and design interventions that survive contact with the actual organization. Those are judgment skills. And those are the skills that experienced agile practitioners already have.
But “having judgment” and “applying judgment consistently under pressure” are different things. You know this from your own practice:
- The Scrum Master who understands systems thinking but still runs Retrospectives on autopilot.
- The Product Owner who knows outcome orientation but still writes feature tickets.
- The Agile Coach who preaches inspect-and-adapt but has not inspected their own career assumptions in three years.
As I described in Part 2 of the “Optimistic” series, most practitioners are in what Virginia Satir called the chaos stage: the old professional identity is cracking, and the new one has not formed yet. [2] That is not a comfortable place to be. But it is also not a place where you can think your way out by reading articles. You need structured practice. You need to run these protocols on real problems, repeatedly, until the thinking pattern becomes yours.
These three skills help close that gap. They impose structure on the moments where judgment matters most: before you propose a solution (Socratic Explorer), after you have committed to a direction (Brutal Critic), and before you execute (Pre-mortem). They do not replace your thinking. They pressure-test it.
Why Agile Practitioners Specifically
If you have spent years in Agile, you already have the raw material. You understand empirical process control (inspect, adapt). You know that retrospection only works if it is honest and that plans fail. The question is how fast you detect and respond.
These three skills operationalize what you already believe. They turn your Agile instincts into repeatable, AI-assisted protocols:
The Socratic Explorer is an instance of inspect-and-adapt applied to problem understanding, not just process adjustment. The Brutal Critic is the Retrospective for your own strategic thinking, with the social pressure removed, so you can confront what is broken. The Pre-mortem is risk management done right: not a risk register that nobody reads, but a failure scenario specific enough to generate a tripwire you can act on.
Applying these skills to your work is how you transition from “the person who runs Scrum events” to “the person the organization turns to when an AI initiative is stuck.” The organizations trying to adopt AI right now need someone who can diagnose why the initiative is failing, stress-test the proposed fix, and identify hidden assumptions before committing more resources. That is what these skills train you to do.
How It Works
The download is a zip file containing three .skill files and an installation guide. You install them in Claude (it takes less than a minute per skill), and they are available to you whenever you need them.
Here is what this looks like in practice. Suppose your organization’s AI pilot has stalled: two teams adopted Copilot three months ago, usage is declining, and leadership is asking you why the promised productivity gains have not materialized. You type: “Socratic: diagnosing why our AI pilot lost momentum after initial adoption.”
The Socratic Explorer does not hand you an answer. It walks you through the problem space first: what makes AI adoption stall in general, what principles distinguish successful adoption from theater, and then applies those principles to your specific pilot, your teams, and your organizational constraints. Twenty minutes later, you have a structured diagnosis and a document you can bring to the next leadership conversation.
Now you have a proposed intervention. Before you pitch it, you type: “brutal critic” and paste your plan. The skill identifies the assumption you did not examine, the cost of what you are avoiding, and what someone who has already turned a stalled pilot around would do differently.
Finally, you type: “pre-mortem,” and the skill assumes your intervention failed in six months. It tells you the three most likely causes and gives you a 30-day tripwire, a specific metric to watch so you know early if the failure is already in motion.
That is three structured thinking sessions, each one building on the last, applied to a real problem you face this quarter: no generic prompts, no “give me 10 ideas,” but applied judgment.
You own these skills. They do not expire. As you use them, you will notice something: the structure they impose starts to shape how you think, even without the tool. That is the real return on this investment.
Finally, don’t forget that inspect and adapt also applies to these skills. The ones you download are just the beginning; iterate on them.
Conclusion
The tool is neutral; your expertise is not. These three AI Thinking Skills do not make you smarter; they make your existing judgment more consistent, more honest, and harder to dodge. Download, install, and use them on a real problem this week. Then iterate.
If in doubt, go back to first principles. These three Claude skills help in the process.
References
[1] Gary Klein: Performing a Project Premortem, Harvard Business Review, September 2007. The underlying research: Deborah J. Mitchell, Jay Russo, and Nancy Pennington (1989).
[2] Virginia Satir, John Banmen, Jane Gerber, and Maria Gomori, The Satir Model: Family Therapy and Beyond (Palo Alto, CA: Science and Behavior Books, 1991). For an accessible summary: Steven M. Smith: The Satir Change Model.