Chapter 3: Prompt design: Linguistic elements

Introduction

The introduction shows how precision, directness, and brevity can explain prompt-output problems even when the task, context, and output format are already specified.

• Three linguistic adjustment dimensions: Explain how precision, directness, and brevity change a prompt's expression and provide distinct ways to diagnose an unsatisfactory output.
  Why it matters: These distinctions help you revise a prompt deliberately instead of changing wording at random.
• CI/CD diagram prompt contrast: Describe why the basic CI/CD analytics diagram prompt leaves more to interpretation than the strategic version, and how their resulting diagrams differ.
  Why it matters: The contrast makes the effects of the three linguistic dimensions visible in one technical task.

Key terms: linguistic elements, precision, directness, brevity, basic approach, strategic approach, CI/CD analytics architecture diagram

3.1 Precision

This section explains how prompt precision shapes output consistency, when to use more or less of it, and what limits its effect.

• Precision and output range: Explain that precision means stating requirements specifically and narrowing how the model should interpret them, which usually narrows the range of plausible outputs and improves consistency. Illustrate the difference using the vague and specific migration-notice prompts.
  Why it matters: This helps you tell whether a prompt gives the model enough direction for the output you need.
• Task fit and precision costs: Justify using higher precision when consistency, accuracy, or system fit matters and lower precision when exploration matters more. Describe the tradeoff: lower precision can increase revision work, while added detail costs tokens and unclear or conflicting requirements can cause filler or failure.
  Why it matters: Choosing precision to fit the task avoids both unnecessary restriction and avoidable rework.
• Other influences on variability: Distinguish prompt precision from other influences on outputs: temperature also affects variability, and clear instructions, specific context, and explicit output format can reinforce precision.
  Why it matters: Output variation should not automatically be attributed to the wording of the prompt alone.

Key terms: precision, plausible outputs, consistency, high precision, lower precision, token cost, sampling settings, temperature, structural elements

3.1.1 Practical example 1: API migration notice

The API migration notice example shows how a precise prompt produces a more actionable engineering notice and can reduce editing time.

• Actionable migration notice specifications: Explain why the precise API migration prompt produces a more usable notice than the vague prompt. Describe how its specified facts, sections, and tone help engineering readers identify the changes, their effects, and required actions.
  Why it matters: A migration notice is useful only if engineers can tell what changes and what they need to do.
• Upfront precision time tradeoff: Explain why spending more time specifying a prompt can reduce the total time spent producing a usable notice.
  Why it matters: Prompt design time can pay off when it avoids substantial revision later.

Key terms: low-precision approach, high-precision approach, section structure, explicit facts, required actions, higher precision, up-front thought, downstream editing

3.1.2 Practical example 2: Sequence diagram generation

A password-reset sequence diagram example shows how prompt precision affects whether the result is a generic sketch or review-ready documentation, and when to increase precision during a project.

• Precision in sequence diagrams: Explain why the low-precision password-reset prompt yields a plausible but incomplete diagram, while the high-precision prompt yields a more review-ready one.
  Why it matters: Specifying what a diagram must cover helps make it useful for review rather than merely plausible.
• Precision across project phases: Describe how to adjust precision as a complex project moves from exploring directions to developing an implementable solution.
  Why it matters: Changing precision with the project phase preserves room to explore before narrowing the work toward execution.

Key terms: low-precision prompt, high-precision prompt, actors, branches, review-ready documentation, moderate precision, two-stage strategy, creative exploration, execution control

3.2 Directness

This section distinguishes directness from precision and explains when implicit or explicit instructions are useful.

• Precision and directness distinction: Distinguish precision, which controls how specific the content is, from directness, which controls whether an instruction states its goal explicitly or leaves it implicit. Explain that directness concerns instruction style, not an aggressive tone.
  Why it matters: Separating these controls helps you identify what to change when a prompt is unclear.
• Implicit and explicit prompt choices: Explain that implicit prompts can encourage conversational exploration of options, while explicit prompts are better suited to responses organized around decisions, ownership, and next actions. Qualify this as a tendency rather than a hard rule, and state why explicit instructions remain the safer default for consistent, execution-ready results.
  Why it matters: Matching instruction style to the desired answer makes a prompt more useful for either exploration or action.

Key terms: precision, directness, implicit prompt, explicit prompt, implicit prompts, explicit prompts, discursive responses, execution-ready results

3.2.1 Practical example 1: Backend API change review

This example contrasts conversational and direct instructions for the same backend API release review.

• Direct versus implicit release instructions: Distinguish the two prompts by explaining that they hold the API change facts and underlying judgment constant, but the explicit prompt directly requests a release decision, required fixes, and immediate next steps.
  Why it matters: The comparison isolates instruction style so you can see what a more direct prompt asks the model to deliver.

Key terms: implicit instruction prompt, explicit instruction prompt, release decision, required fixes, immediate next steps

3.2.2 Practical example 2: Flaky CI test strategy

A flaky-CI example shows how an implicit request tends to invite discussion, while an explicit instruction asks for a decision and rollout plan, without guaranteeing either response style.

• Directness in CI recommendations: Distinguish the implicit request to share thoughts on three flaky-test approaches from the explicit instruction to choose one approach and give a decision, risk, and first rollout step. Describe how the example responses differ.
  Why it matters: Direct wording helps you request a decision rather than just an exploration of options.
• Limits of directness: Explain why the contrast between these two responses is directional rather than guaranteed.
  Why it matters: Direct wording steers a response, but it does not guarantee certainty or prevent other useful output.

Key terms: implicit instruction prompt, explicit instruction prompt, directness, test isolation, directional, not deterministic, implicit prompts, explicit prompts

3.3 Brevity

This section explains how to keep prompts no longer than needed while retaining detail that improves the result.

• Task-appropriate prompt length: Explain how to choose prompt length based on task complexity: keep routine requests brief, but add concrete constraints or success criteria when specialized or high-stakes work needs them.
  Why it matters: This helps you avoid both vague short prompts and long prompts that add no useful direction.
• Filler and token cost: Explain why removing repeated politeness and restated instructions can improve prompt efficiency without removing clear requirements.
  Why it matters: Removing filler saves tokens while preserving the requirements that guide the output.

Key terms: brevity, minimum effective prompt length, constraints, success criteria, input token, filler, uncached prompt tokens

3.3.1 Practical example 1: Incident summary analysis

The section compares two prompts for summarizing on-call notes to show how a shorter, focused request can preserve the needed incident details without inviting unnecessary output.

• Strategically brief incident requests: Explain how the shorter incident-summary prompt states the needed scope without the longer prompt’s extra wording. Identify root cause, timeline, and customer effect as the requested elements.
  Why it matters: A concise prompt can still make the required content explicit.
• Cost of excess framing: Explain why the longer prompt is inefficient for the immediate goal and why that inefficiency matters in repeated incident-summary workflows.
  Why it matters: Small amounts of avoidable prompt text become more consequential when a team processes many incidents.

Key terms: strategically brief, root cause, timeline, customer effect, prompt cost, remediation, follow-up detail, prompt efficiency

3.3.2 Practical example 2: Technical diagram generation

This section compares two telemetry-diagram prompts to show how strategic brevity preserves essential structure while targeted details add scope when needed.

• Concrete diagram requirements: Explain why the brief telemetry-diagram prompt retains the essential pipeline structure, and distinguish its concrete components and flows from the longer prompt’s generic polish language and additional failure-handling requirement.
  Why it matters: It helps you spend prompt space on diagram requirements rather than descriptions of how polished the result should feel.
• Requirement driven brevity: Describe when to keep a diagram prompt brief and when to add detail, using the section’s advice to add targeted specifications for required features or observed gaps.
  Why it matters: This gives you a way to add useful detail without padding every prompt by default.

Key terms: strategically brief, concrete components and flows, generic polish language, failure handling, targeted specifications, task requirements, model’s baseline capability, observed gaps

3.4 Additional linguistic considerations

The section explains when secondary wording choices help align a prompt, while keeping precision, directness, and brevity as the primary controls.

• Primary and secondary controls: Distinguish precision, directness, and brevity from the secondary linguistic refinements described here, and explain why the refinements may still be useful.
  Why it matters: This helps you prioritize the prompt changes most likely to affect core behavior.
• Positive and visual specifications: Explain why the section favors positive specifications over negative-heavy restrictions, especially when requesting technical visuals.
  Why it matters: Concrete requests give the model a clearer target than a list of things to avoid.
• Audience-matched tone guidance: Describe when explicit tone guidance improves a prompt and what it should align with.
  Why it matters: Explicit tone guidance can make outputs consistent when the expected voice is not obvious.
• Emphasis word effects: Explain how urgency or risk words can affect a model’s response and when to avoid them.
  Why it matters: A loaded adjective can steer attention even when the underlying task has not changed.

Key terms: precision, directness, brevity, secondary refinements, positive specifications, negative instructions, visual content framing, components, flows, boundaries

3.4.1 Question vs. statement formatting

This section explains when question or statement formatting may help and why it should be a final refinement rather than a primary prompt control.

• Question and statement fit: Distinguish when question formatting may suit brainstorming or research inquiries and when statement formatting may suit implementation planning or decision support.
  Why it matters: Choosing a form that fits the task can help align the response with what you need.
• Formatting as final refinement: Explain when to adjust question-versus-statement formatting relative to precision, directness, and brevity.
  Why it matters: It keeps a formatting change from standing in for clearer prompt design.

Key terms: questions, statements, brainstorming, implementation planning, decision support, refinements, primary controls, precision, directness, brevity

3.4.2 Hands-on practice

This hands-on section turns precision, directness, and brevity into deliberate choices by having readers compare prompt variants on realistic engineering tasks.

• Controlled linguistic prompt iteration: Apply precision, directness, and brevity deliberately by writing variants for the same engineering task, comparing their outputs, and identifying which wording reduces revision work without losing useful flexibility.
  Why it matters: Comparing purposeful variants helps you improve a prompt without rewriting it at random.

Key terms: precision, directness, brevity, prompt variants, controlled design process

3.5 Summary

This summary explains how deliberate language choices can reduce prompt revision cycles while balancing execution, exploration, cost, and communication consistency.