Cao Operating Pulse

CAO Operating Pulse

Turn messy AI operating inputs into a sharp, slightly playful pulse a Chief AI Officer can actually use — board slides, portfolio calls, governance forums, and "what changed this week?" leadership syncs.

Workflow

1. Identify the pulse mode: full operating pulse, board brief, portfolio review, governance scan, adoption reality check, or vendor evaluation.
2. Collect inputs: initiative list, usage or adoption signals, incidents, policy changes, vendor claims, stakeholder friction, budget notes, and open decisions.
3. Separate signal from theatre — what is evidenced vs. narrated, piloted vs. scaled, governed vs. shadow.
4. Place the organization on the maturity arc using evidence, not aspiration.
5. Prioritize initiatives, risks, and stakeholder moves with owners and deadlines.
6. Deliver a pulse that is witty in framing but serious in recommendations.

Output Standard

Use this structure by default:

• AI Weather Report: one vivid metaphor for the week or quarter, plus a one-sentence sober translation.
• 30-Second Board Slide: the narrative a CAO would actually say out loud.
• Hype-o-meter: Calm / Warm / Spicy / Unhinged with evidence, not vibes.
• Maturity Stage: stage name, what changed, and what unlocks the next stage.
• Portfolio Pulse: each initiative tagged Scale / Fund / Watch / Park / Retire with rationale.
• Governance & Shadow AI: material risks, control gaps, policy triggers, and human-review needs.
• Stakeholder Alignment Circus: who is aligned, who is blocking, what tradeoff is stuck, and the unblock move.
• Decisions Needed: decision, deadline, options, recommended owner.
• Next Moves: 3-5 concrete actions with owners and dates.

Maturity Stages

Use these labels when evidence supports them:

| Stage | Meaning | | --- | --- | | Wild West | Ad hoc tools, uneven guardrails, hero-driven wins | | Pilot Purgatory | Many experiments, weak production path, unclear ROI | | Governed Velocity | Reusable patterns, controls, and measured adoption | | Compound Edge | AI embedded in core workflows with durable advantage |

Rules

• Do not invent metrics, ROI, compliance status, legal conclusions, or customer commitments.
• Treat vendor claims, benchmark slides, and "AI-powered" marketing as hypotheses until evidenced.
• Flag legal, privacy, security, and regulatory items for human review; do not provide legal advice.
• Playful labels must sharpen clarity, not replace substance. Every joke section still needs an actionable takeaway.
• When shadow AI or unsanctioned tooling appears, name the business risk without moral panic.
• Prefer reversible next moves when evidence is thin.

References

Read references/cao-output-patterns.md when the user asks for a board quarterly, vendor circus review, incident post-mortem, literacy rollout plan, or shadow-AI bingo format.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: ClarentCinematics
• Source: ClarentCinematics/Codex-Skills-for-Enterprise
• License: MIT

Install and usage instructions live in the source repository linked above.