Break down tasks: workflow thinking

Break down tasks: workflow thinking

Why it matters: Breaking work into steps makes AI outputs controllable and reduces errors.

Goals

• Decompose a task into a workflow.
• Write prompts per step.
• Add a QA step at the end.

Key definitions

• Workflow: A sequence of steps that turns inputs into outputs with checks.
• Input contract: What information must be provided (fields, constraints, examples).
• Output contract: Exact format and criteria the output must satisfy.
• QA step: A final check that tests for correctness and completeness.

Workflow (step-by-step)

1. Define the final deliverable (format + audience).
2. List 3–6 steps: gather inputs → draft → refine → format → QA.
3. Write one prompt per step (include input/output contracts).
4. Add a QA prompt that checks against criteria and flags uncertainties.
5. Run the workflow end-to-end once and refine the weakest step.

Example (good vs bad)

✅ Good: A multi-step workflow produces consistent outputs and catches errors before shipping.

❌ Bad: A single mega-prompt tries to do everything and fails unpredictably.

Checklist

• Steps are numbered.
• Each step has clear inputs/outputs.
• QA checks facts + formatting.
• Workflow is reusable.

Metrics / criteria

• Error rate drops over time as you refine the weakest step.
• Measurable: You can say whether the output is correct/complete (pass/fail or a score).
• Effort: You can produce the result in 15–30 minutes using the workflow.
• Clarity: Another person can run your prompt and get a similar outcome.

Common mistakes + fixes

• Pitfall: Too many steps. Fix: Start with 3–5 steps; add only when necessary.
• Pitfall: No QA step. Fix: Always include a verification prompt before shipping.
• Pitfall: Vague instructions. Fix: Add constraints, examples, and a success checklist.
• Pitfall: One-shot prompting. Fix: Iterate: draft → critique → revise → verify.
• Pitfall: No validation. Fix: Add QA steps, tests, and explicit “what counts as done”.

Student tasks

1. Apply the workflow to one real work task related to break down tasks: workflow thinking.
2. Create a small artifact: prompt, checklist, rubric, table, draft, or decision note.
3. Run one quality check and record what changed after the check.
4. Write a 3-sentence reflection: what worked, what failed, and what you will reuse.

Useful external sources

• OpenAI prompt engineering guide: https://platform.openai.com/docs/guides/prompt-engineering — Prompt structure, iteration, constraints, and evaluation habits.
• NIST AI Risk Management Framework: https://www.nist.gov/itl/ai-risk-management-framework — Risk framing, governance language, and verification mindset.
• OWASP Top 10 for Large Language Model Applications: https://owasp.org/www-project-top-10-for-large-language-model-applications/ — Practical AI safety risks such as prompt injection, leakage, and insecure output handling.
• Google People + AI Guidebook: https://pair.withgoogle.com/guidebook/ — Human-centered AI design, user control, feedback, and failure modes.

Bibliography

• OpenAI prompt engineering guide. https://platform.openai.com/docs/guides/prompt-engineering
• NIST AI Risk Management Framework. https://www.nist.gov/itl/ai-risk-management-framework
• OWASP Top 10 for Large Language Model Applications. https://owasp.org/www-project-top-10-for-large-language-model-applications/
• Google People + AI Guidebook. https://pair.withgoogle.com/guidebook/