Playbooks · how to actually use AI

The AI playbooks.

Concrete, copy-and-adapt playbooks for getting real work out of AI — privately. Start with the one skill that changes everything (a good prompt), learn to give the AI lasting knowledge, then climb from assisted drafting to supervising agents. Real examples throughout.

Playbook 01 · the prompt

A good prompt is the whole game.

Most “the AI got it wrong” moments are really “the AI wasn’t told enough.” Same model, two prompts — watch what changes.

✕ Vague
Write a board report for our association.

Generic, hedged, and full of invented numbers — because you handed it nothing to work with.

✓ Specific
You are a senior ops manager. Draft the Q3 board financial summary from the data below, using only these figures…

Accurate, on-format, and ready to review. The full version has six parts ↓

The anatomy of the specific prompt:

ROLEYou are a senior association operations manager preparing a quarterly board packet. TASKDraft the Q3 financial summary section of the packet. CONTEXTUse this data: the Q3 P&L from Sage Intacct, membership counts from Salesforce NPSP, and event attendance from Cvent. {{paste scrubbed data}} CONSTRAINTSUse only the figures provided — never estimate or invent a number. Flag anything that doesn’t reconcile across the three sources. One page maximum. FORMATLead with a 3-sentence executive summary, then a metrics table with year-over-year change, then a short narrative on the biggest variances. EXAMPLESMatch the structure and tone of last quarter’s approved summary. {{paste example}}

Rule of thumb: before you blame the model, add the missing part — usually context or a constraint. When prompting isn’t enough: RAG vs. fine-tuning →

Playbook 02 · artifacts & memory

Stop re-explaining yourself.

A model starts every chat blank. Artifacts are the durable knowledge it reuses; memory is the context it carries between sessions. Capture these once and every future run gets better. Real examples of artifacts worth keeping:

Artifact · style guide

Voice: plain and direct, no jargon. Spell out every acronym on first use. Currency as $12,400 — no cents in summaries. Never use the word “synergy.”

Artifact · approved example

Last quarter’s board summary is the gold standard for structure, length, and tone. Every new draft should look like this.

Memory · terminology

A “member” is active and dues-paying; “lapsed” means 90+ days overdue; “the board” meets quarterly. Remember these across every task.

Artifact · SOP / checklist

Board-packet steps: 1) pull P&L, membership, attendance 2) reconcile across sources 3) flag mismatches 4) draft summary 5) route to the director for approval.

Artifacts + memory turn a generic model into one that knows your organization — your own knowledge base, compounding with every run.

Playbook 03 · climbing the ladder

What to actually do at each rung.

Power with AI is earned one rung at a time — and a human stays in control the whole way. Here’s the move at each step, not just the name.

RUNG 01 · Assisted

Draft & learn

  • Write every prompt with the six parts above.
  • Paste in real context — scrubbed of anything sensitive.
  • Save your best outputs as artifacts.
  • Verify every number and claim before you trust it.
RUNG 02 · Supervised

Do one real task

  • Connect one tool and run one real task end to end.
  • Start with low-risk, reversible work.
  • Approve every external action yourself.
  • Read the audit trail to learn how it behaves.
RUNG 03 · Workflows

Make it repeatable

  • Take a run that worked twice and capture it.
  • Save its prompt + artifacts as a reusable workflow.
  • Add memory for the recurring context.
  • Schedule it; switch to approve-by-exception.
RUNG 04 · Autonomous

Supervise agents

  • Scope a clear goal and firm guardrails.
  • Let a long-horizon agent run for hours or days.
  • Supervise via the exception queue and audit trail.
  • Review outcomes, not every step — keep an escalation path.

See the interactive ladder →

Playbook 04 · keep it private

The do / don’t checklist.

All the leverage, none of the leakage. Pin this up.

Use a governed gateway or a no-train tier with a signed DPA.
Strip PII before anything leaves your boundary.
Keep API keys server-side — scoped, rotated, never in client code or git.
Run open models in your own environment for crown-jewel data.
Never paste donor lists, contracts, or financials into a consumer chatbot.
Never put a key in client-side code or a git repo — bots scrape new commits in minutes.
Never trust a number you can’t trace back to a source.

The specific risks, in depth: how your data ends up in training sets, and how keys leak →

We run these for you

Want your team running these playbooks?

Enablement walks your people up every rung — by role, on your real work, governed at each step. No engineering background required.