# Waymaker Agency — AI workflow practice pack

Published 2026-10-16. Original educational prompts and exercises. Use approved, minimized source material. Outputs need human review.

## Build an AI client-reporting workflow with a human approval gate

Create a repeatable reporting process that helps an agency explain performance clearly while keeping numbers, recommendations, and delivery under review.

### Draft a narrative that exposes uncertainty

Draft a client performance brief from this verified table and context: [table, definitions, dates, notes]. Separate observed changes, plausible explanations, and proposed next actions. Cite the table row or source reference for every numeric claim. Mark missing or conflicting information. Do not claim causation from correlation, invent benchmarks, promise outcomes, or recommend budget changes as already approved. End with the decision we need from the client.

### Test the unpleasant cases

Audit this draft against the supplied source table, client instructions, and approval criteria: [materials]. List each numeric inconsistency, unsupported explanation, ambiguous comparison, accidental commitment, and missing caveat. Show the exact statement and supporting or conflicting evidence. Return ready for human approval or needs revision. Your response is a review aid and must not trigger delivery.

### Your exercise

1. Choose one existing client report and document its decision and metric definitions.
2. Generate a draft from a verified sample, then complete both review passes.
3. Test a rejected approval and a missing source before enabling any recurring run.

### Review checklist

- [ ] Every number traces to verified data.
- [ ] Observed changes and explanations are distinguished.
- [ ] Approval applies to the exact report version.
- [ ] No scheduler or delivery step is assumed to exist because a prompt mentions it.

Your evidence:

Your next decision:

Owner and review date:

## Turn your best repeatable service into an AI-assisted delivery playbook

Document the decisions, evidence, and quality standards behind a service so AI helps your team deliver it consistently.

### Assign a useful role to AI at each stage

Convert these redacted examples and working notes into a draft service playbook: [materials]. For each stage, define required inputs, tasks AI can assist, human decisions, expected output, and a review standard. Identify gaps instead of filling them with invented process. Separate client-specific choices from reusable methods. Keep the playbook focused on this deliverable: [deliverable].

### Run a reviewer comparison before expanding

Review this deliverable against our brief, evidence, and quality rubric: [materials]. For each criterion, return pass, revise, or cannot assess, with the exact evidence. Identify unsupported claims, missing customer context, and recommendations that exceed the agreed scope. Suggest the smallest changes needed for approval. Do not lower the standard to make the draft pass.

### Your exercise

1. Select one repeatable deliverable and gather three permitted examples.
2. Document the five stages and identify one useful AI task per stage.
3. Test the rubric on a draft with a deliberately unsupported finding.

### Review checklist

- [ ] The service has a concrete definition of done.
- [ ] AI assistance and human decisions are assigned explicitly.
- [ ] Examples are permitted, redacted, and labeled where fictional.
- [ ] The playbook has an owner, version, and maintenance process.

Your evidence:

Your next decision:

Owner and review date:

## Keep learning

Free beginner course: https://aiteachers.cc/courses/first-useful-ai-workflow/

Free experienced-user course: https://aiteachers.cc/courses/repeatable-ai-workflows/