Support

Revising a Support Prompt Safely

Support prompts carry policy. A worked revision showing how to evolve one without dropping the boundary that keeps replies inside policy.

Open in Prompt Version Diff

Overview

A customer support prompt is operational policy wearing a prompt's clothes: the refund boundary, the escalation rule, the promise discipline. That makes revisions higher-stakes than they look — an edit that improves warmth while loosening the boundary is a net regression with a friendly face. This resource diffs a realistic support prompt revision and shows the safe pattern: tone and structure can move freely, policy lines must appear unchanged or strengthened in the diff, never in the removed list.

How to use this resource

Diff the loaded revision

Version B reworks tone and adds structure. Run the diff and go straight to the removed list — it should be empty of policy lines.
Verify the boundaries

Both policy lines survive in the modified list, strengthened: verification got a timeline, escalation got a third trigger.
Check what tone edits added

'No generic empathy lines' is a constraint, not fluff — it shows as added control, not added ambiguity.
Apply the rule

For support prompts: policy lines may appear in added or modified, never in removed. One glance at the diff enforces it.

Why This Works

Support prompts fail on policy before they fail on tone — the diff checks policy first
Modified-list review shows whether a boundary was strengthened or quietly relaxed
The never-in-removed rule is simple enough for every editor to apply consistently

Best for

Support teams running shared reply prompts
Prompts where a loosened boundary creates real financial exposure
Revisions that touch both tone and policy in one pass

Not for

Choosing between two support prompt drafts — that's the Prompt Comparator
Compressing a bloated support prompt — that's the Prompt Cleaner

Use cases

Reviewing edits to a support prompt that encodes refund or escalation policy
Onboarding a new support lead onto a prompt's policy lines
Confirming a tone rewrite left every boundary intact

FAQ

How do I check a support prompt revision didn't weaken a policy boundary?

Run the diff and go straight to the removed list — the rule is that policy lines like "Refunds only after the duplicate charge is confirmed" and the chargeback/legal escalation trigger may appear in added or modified, never in removed. In Version B both survive strengthened. The prompt-version-diff shows the movement; confirming the policy intent held is your review, not a compliance guarantee.

Is adding a rule like 'no generic empathy lines' a tone change or a constraint?

This resource treats "no generic empathy lines" as added control, not added ambiguity — a constraint that shows up in the added list rather than fluff. It illustrates the safe pattern: tone and structure can move freely in a revision as long as the diff shows policy lines strengthened or intact. Whether the reworked wording actually stays inside policy is still a human judgment.

Customize This Resource

Opens both versions in Prompt Version Diff. Run the diff to see what changed and whether the revision improved the prompt.

Open in Prompt Version Diff

Version A (v1)

Copy it as-is, or use Open in Prompt Version Diff to load it pre-filled and customize it with your own context.

Reply to customers reporting billing issues.
Be empathetic and professional.
Refunds only after the duplicate charge is confirmed in the billing system.
Escalate if the customer mentions a chargeback or legal action.

Version B (v2)

Reply to customers reporting billing issues.
Open by acknowledging the specific problem in one sentence — no generic empathy lines.
Refunds only after the duplicate charge is confirmed in the billing system; say what verification involves and how long it takes.
Escalate if the customer mentions a chargeback or legal action, or has contacted us three or more times about the same charge.
Tone: calm and specific. Under 150 words, end with the next step.

More resources from Prompt Version Diff

Resource

Compare Two Versions of a Prompt

See exactly what changed between v1 and v2 of a prompt — added, removed, and modified instructions, plus whether the revision reduced or introduced risk.

Prompt Engineering

Resource

Track Prompt Changes Across Revisions

A lightweight way to track how a prompt changes over time: keep the previous version, diff every revision, and read the risk deltas instead of guessing.

Prompt Engineering

Resource

Revising a Marketing Prompt: v1 to v2

A worked marketing prompt revision — from adjective soup to offer-driven — with the diff showing exactly which changes carry the improvement.

Content

Resources that pair well

Resource

Compare Two ChatGPT Prompts

A side-by-side way to decide between two ChatGPT prompt drafts — scored on clarity, specificity, output control, and risk instead of gut feeling.

Prompt Engineering

Resource

Prompt Cleanup Examples (Before & After)

A set of before-and-after examples showing exactly what prompt cleanup removes — and what it deliberately leaves alone.

Prompt Engineering

Resource

Agent Instruction Prompt Formatter

Formats fuzzy agent instructions into a structured prompt with objective, available tools, constraints, success criteria, and failure handling.

AI Agents

Related tools

Tool

Prompt Version Diff

Diff two versions of the same prompt — added, removed, and modified instructions, plus risk changes.

Prompt Builders

Tip: Save time by exploring related resources and tools that integrate with this resource.