Skills that help senior executives think clearly about AI
SaferSkills independently audited boardroom-ai-skills (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
There is a gap between what executives are being told about AI and what they can actually verify.
Vendors demo capability that doesn't survive contact with operations. Consultants publish adoption frameworks that describe a different organisation than the one you run. Boards ask questions nobody has clean answers to. And somewhere in the middle, senior executives are expected to have a position — on timelines, on investment, on risk — that they were never equipped to form.
This repo is for that gap.
These are skills for Claude — small, precise prompts that help you think through the decisions AI is forcing on you. They are not productivity tools. They are epistemic tools: designed for a moment when the hardest part of your job is knowing what to believe, who to trust, and what questions you're not being asked.
The result is a specific kind of confusion that is hard to name out loud:
to know what skepticism is warranted
that means operationally
readiness — and you can't tell which is which
your organisation's actual capacity to absorb change
These skills are for the executive who is frustrated by that confusion. They won't tell you to adopt AI faster. Some of them will tell you why you're moving too fast. All of them will tell you what is actually in front of you.
That is what most AI advice for executives is not designed to do.
These skills work inside Claude. To use any skill:
SKILL.md fileNo setup required. No data leaves your conversation. You don't need a technical background — you need the problem the skill is designed for.
Before strategy, there is the moment of pressure: the board question you weren't ready for, the vendor demo that felt impressive but unverifiable, the internal divide you can't quite name. These skills are for that moment.
| Skill | What it does |
|---|---|
| name-the-pressure | Surface the exact form your AI adoption pressure is taking — board mandate, competitive signal, internal fracture, vendor migration, talent pressure — before you're required to respond publicly. Each type requires a different response. Most executives are responding to all of them at once without naming them separately. |
| ai-theatre-detector | After a vendor pitch, internal proposal, or board presentation: run a structured diagnostic across six dimensions to distinguish genuine AI capability from performance. Produces a calibrated verdict and the one question to ask before any further engagement. |
| vendor-interrogation | Generate the questions that cut through the demo. Most AI vendor meetings are structured to answer questions you didn't ask. This skill generates the ones you should have asked — organised by the moment in the meeting when they land hardest. |
| Skill | What it does |
|---|---|
| board-narrative | Turn your current AI posture into a coherent narrative for a board or investor audience. Structured around four distinct narrative postures — Deliberate Progression, Informed Caution, Active Reorientation, Foundation Building — so you're not inventing the frame from scratch. |
| ai-strategy-brief | Produce a one-page internal brief using a three-column structure: what you're doing, what you've decided not to do, and what you're deferring and why. The discipline of completing all three columns is the strategy. |
| build-buy-partner | Structured framework for deciding whether to build AI capability internally, buy it from a vendor, or access it through a partnership — calibrated to the specific dynamics of AI: rapid depreciation, vendor lock-in risk, talent scarcity, and the difference between commodity and strategic capability. |
| Skill | What it does |
|---|---|
| ai-readiness-read | Four-dimension assessment that surfaces your organisation's actual AI readiness — separate from its stated readiness. Identifies genuine capability pockets, maps the priority gaps, and produces the one thing that, if fixed, would most improve your position in the next 6 months. |
| change-resistance-map | Map the resistance landscape before an AI initiative is announced. Five resistance types — threat, trust, workload, governance, coordination — each with specific signals and specific responses. Most AI initiatives fail here, not in the technology. |
| Skill | What it does |
|---|---|
| powertalk-ai | Five specific moves that shift AI conversation from posture to substance: the honest anchor, the specific claim, the named uncertainty, the calibrated disagreement, and the honest "I don't know." For any high-stakes room where generic AI confidence is the norm and calibrated clarity is the differentiator. |
| stakeholder-translation | Translate the same AI position into the specific register of each audience — board, investors, regulators, clients, employees, media — without changing the underlying substance or creating contradictions across versions. |
I build infrastructure for AI systems — the layer below the demos, where the real constraints live. I've designed cryptographic protocols for AI agent identity, geospatial intelligence platforms, and multi-agent architectures. My orientation is toward the foundational layer: what has to be true for the thing above it to work.
I built these skills because the executives I talk to are being handed the wrong tools for the problem they have. The problem is not AI literacy. It is epistemic navigation — knowing what to believe, in a domain where the incentives of almost every person in the room are misaligned with telling you the truth simply.
If you want to think through what you're navigating — not a pitch, not a proposal, just a conversation — I'm occasionally available for exactly that.
If you've used these skills and found them useful — or found them wrong — I want to hear from you. The skills improve with contact with the actual problem.
Open an issue or reach out directly.
MIT License. Use freely. Attribution appreciated but not required.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.