Create a copy-paste prompt library for an operations team at [COMPANY OR TEAM]. Our operating context is [OPERATING CONTEXT], and our recurring work includes [RECURRING WORK]. The people using these prompts are [USERS], and the systems or source materials they can provide are [SYSTEMS AND DATA]. Our main operational risks and constraints are [RISKS AND CONSTRAINTS]. Produce 10 distinct prompts covering the most relevant mix of: weekly operating review; SOP drafting or repair; process mapping; capacity and staffing planning; KPI definition; project handoff; vendor evaluation; root-cause analysis; incident communications; and decision logs. Skip topics that do not fit our context and replace them with a more relevant recurring task. For each prompt, provide: a clear title; when to use it; 3 to 6 bracketed inputs; the complete copy-paste prompt; and a one-line check the user should make before acting on the result. Each prompt must request a concrete operational artifact, such as a RACI table, swimlane process map, metric dictionary, experiment backlog, escalation matrix, or corrective-action register. Require assumptions, dependencies, owners, dates, and measurable completion criteria where appropriate.
Build in operations discipline: distinguish a symptom from a root cause; identify process handoffs and exception paths; avoid assigning an owner without authority; and do not treat a dashboard metric as proof of causation. Do not invent system data, policy, staffing availability, vendor terms, or compliance requirements. Ask up to 3 clarifying questions only if a required input is missing. Before answering, check that every prompt fits a real recurring decision or workflow and that its requested artifact can be completed from the inputs it asks for.
Fill in
| Placeholder | What to enter | Example |
|---|---|---|
| [COMPANY OR TEAM] | Enter the team name, business type, and size. | Northline Fulfillment, a 22-person e-commerce operations team |
| [OPERATING CONTEXT] | Enter the product or service, locations, volume, and operating model. | Two warehouses ship 1,500 orders per weekday for home-goods brands; work runs in two shifts. |
| [RECURRING WORK] | Enter the workflows, meetings, and decisions the team repeats. | Daily backlog review, pick-pack process changes, seasonal staffing, carrier issues, and client escalations. |
| [USERS] | Enter the roles that will use the prompts. | Warehouse manager, shift leads, client-success manager, and operations analyst |
| [SYSTEMS AND DATA] | Enter the dashboards, tickets, documents, and tools users can supply. | WMS exports, Zendesk tickets, carrier scorecards, staffing roster, and Google Sheets |
| [RISKS AND CONSTRAINTS] | Enter service levels, regulations, budgets, staffing limits, and recurring failure points. | Same-day cutoff at 2 p.m.; mis-ship rate below 0.3%; no additional full-time hires this quarter. |
How to use
- Describe the work volume, handoffs, and available source systems so the library matches actual operations.
- Copy one prompt into your AI tool and attach the report, ticket export, or meeting notes it requests.
- Check owners, dates, system assumptions, and exception paths against the people who do the work.
- Follow up with: “Convert the approved corrective actions into a weekly tracker with owner, due date, evidence, and escalation trigger.”
Variations
SOP builder
Use this to document a process that new hires or cross-functional teams must follow.
Create an SOP for [PROCESS] at [TEAM]. Inputs are [CURRENT NOTES], systems are [SYSTEMS], and the required service level is [SERVICE LEVEL]. Produce purpose, scope, roles, prerequisites, numbered steps, decision points, exception handling, quality checks, escalation contacts, and revision history. Mark gaps instead of inventing policies or system actions. Include a one-page checklist a new operator can use.
Incident review
Use this after a service failure, outage, delay, or quality escape.
Using this incident record [INCIDENT FACTS], prepare a blameless operations review for [AUDIENCE]. Produce timeline, customer or business impact, containment actions, contributing conditions, root-cause hypotheses, evidence needed to confirm each hypothesis, corrective actions, owners, due dates, and effectiveness measures. Separate facts from assumptions. Do not assign personal fault or claim a root cause without evidence.
Capacity plan
Use this before a launch, seasonal peak, or workload change.
Build a capacity plan for [WORKFLOW] over [TIME PERIOD]. Demand forecast is [DEMAND DATA], current capacity is [CAPACITY DATA], staffing constraints are [CONSTRAINTS], and service target is [SERVICE TARGET]. Return demand assumptions, workload units, capacity calculation, bottleneck analysis, scenario table, staffing or scheduling actions, and trigger thresholds. Flag data needed for a defensible forecast; do not invent productivity rates.
Tips
- Write SOPs around decisions and exceptions, not just the happy path; the exception route is where most operational variation happens.
- Name a metric's numerator, denominator, timing, population, and owner before using it in a performance conversation.
- For incident reviews, preserve the timeline before proposing causes so hindsight does not reshape what people knew at the time.
- Use a single corrective-action register with due date and evidence of effectiveness; a completed task is not proof that the problem is resolved.
FAQ
What operations tasks are good for AI prompts?
AI is useful for turning supplied notes and data into SOPs, meeting structures, decision logs, incident reviews, and capacity scenarios. It should not replace access-controlled system actions or accountable operational decisions.
Can AI create an SOP from a rough process?
Yes, but have the people performing the work validate every step, exception, and system permission. A polished document can still describe the wrong process.
How do I stop AI from inventing operational details?
State the source boundary, require an assumptions list, and ask it to label missing facts as questions or gaps.