The prompt for every job the plugin does.

Use the ContextHint plugin to explain what a context hint is in ChatGPT Ads, how it differs from keyword targeting, and how to write a good one. Then walk me through a good example and a bad one, and tell me exactly what separates them.

Use the ContextHint plugin to explain how ChatGPT ad campaigns are actually structured: the ad-group maturity ladder, how to split one product into ad groups, intent laddering, message match, and how specific a hint should be. Then turn it into a checklist I can apply to [PRODUCT OR WEBSITE].

Use the ContextHint plugin to tell me whether ChatGPT Ads is a fit for [PRODUCT OR WEBSITE]. Check the category fit for this kind of product, find the niches with captured ad evidence closest to it, and show me the buyer-intent split there. Give me a go or no-go, and say plainly where the evidence is thin instead of filling the gap.

I run marketing at [BRAND]. Use the ContextHint plugin to find which advertisers on ChatGPT are closest to us. Our brief: [ONE PARAGRAPH ON WHAT THE PRODUCT DOES, WHO IT IS FOR, AND HOW IT DIFFERS]. Rank them, skip the domain aliases, and separate the direct rivals from the adjacent ones. Link every advertiser so I can inspect their captured ads.

Use the ContextHint plugin to break down [RIVAL 1], [RIVAL 2] and [RIVAL 3]: which ad groups each one runs (niche, audience, intent), what their context hints look like, and the actual ad copy running against each ad group. Read the hints one at a time and tell me which buyer each ad group is aimed at. Keep what was observed in captured ads separate from what you inferred.

Use the ContextHint plugin to analyse the [NICHE] market: who advertises most, the buyer-intent distribution, the recurring targeting patterns across advertisers, and where the whitespace is. If the niche turns out to be a broad bucket that mixes different buyers, say so and run a similarity search on my brief instead of treating the sample as my peer set. My brief: [ONE PARAGRAPH].

Use the ContextHint plugin to show me what [MY BRAND] is already running on ChatGPT: the niches and buyer conversations our ads land in, the inferred hint behind each ad group, and our actual ad copy. Then tell me which of those conversations we genuinely want and which look accidental. If we are not in the captured data at all, say so plainly rather than inferring.

Use the ContextHint plugin to find the right niche for [PRODUCT OR WEBSITE]. Search the niche catalogue for [CATEGORY, AUDIENCE OR PROBLEM], then prepare the evidence for the best candidates and read the actual buyer prompts back to me. Reject any niche whose prompts describe a different buyer even when the wording looks similar, and tell me if no niche is a clean fit.

Use the ContextHint plugin to structure a ChatGPT Ads campaign for [PRODUCT OR WEBSITE]. Start from the campaign-structure playbook and the targeting patterns in [NICHE], then split us into ad groups by buyer rather than by feature. For each ad group give me the audience, the buying moment, the intent, and the captured evidence it rests on. Do not over-segment: tell me which ideas should stay merged because the evidence is too thin to separate them.

Use the ContextHint plugin to build [BRAND]'s ad groups out of what the rivals already run. Find the advertisers closest to our brief, read their ad groups and context hints, cluster them into recurring plays, then propose [NUMBER] ad groups for us: the plays worth matching and the buyers nobody is covering. Our brief: [ONE PARAGRAPH].

Use the ContextHint plugin to write and score one context hint for each of these ad groups: [AD GROUP 1], [AD GROUP 2], [AD GROUP 3]. Prepare evidence per ad group, write one candidate each, evaluate them all, and revise once where the evaluation asks. Then rank them and tell me which to run first, which to defer, and which to drop, using the evaluation signals rather than your own impression. Product: [ONE PARAGRAPH].

Here is a context hint we run for [PRODUCT]: [PASTE THE HINT]. Use the ContextHint plugin to audit it. Prepare the evidence, evaluate the hint exactly as written, and tell me whether it scores better than a plain paraphrase of our product description. Then write one revision, evaluate the original and the revision together, and keep whichever actually scored higher. Show me both numbers and do not assume the revision won.

Use the ContextHint plugin to show me the actual ad copy [RIVAL 1], [RIVAL 2] and [RIVAL 3] run, grouped by the ad group and audience each ran against. Point out the patterns: how the titles name the buyer's situation and how the bodies continue the answer. Do not describe any of them as winning or best performing, because the data does not show performance.

Use the ContextHint plugin to draft three title and body pairs for each ad group we are running: [AD GROUP 1], [AD GROUP 2]. Pull the message-match guidance and the real creatives that ran against similar audiences first, then write ours in the same register without copying anyone's language. Keep the titles short and outcome first, write each body as a continuation of the answer the buyer is reading, and label every line you wrote as recommended.

Review this ChatGPT ad against the ad group it is meant for. Creative: [PASTE THE TITLE AND BODY]. Context hint: [PASTE THE HINT]. Use the ContextHint plugin for the message-match rules and the creatives that ran against this audience, then tell me where the message and the conversation diverge, which conversations this would wrongly attract, and what to change. Keep the review separate from any claim about how it will perform.

We have [$AMOUNT] for [NUMBER] days. Use the ContextHint plugin to split it across the ad groups we decided to run: [AD GROUP 1], [AD GROUP 2], [AD GROUP 3]. Only count an ad group as kept if I said it is genuinely our buyer. Show me the structure, the split, the reserve, how many creatives each ad group needs, and which assumptions are illustrative rather than measured. State the limits of the plan before the numbers.

Use the ContextHint plugin to compare a [$LOWER AMOUNT] plan and a [$HIGHER AMOUNT] plan over [NUMBER] days for these ad groups: [AD GROUP 1], [AD GROUP 2], [AD GROUP 3]. Run both, put them side by side, and tell me what the smaller budget has to give up, which single ad group it should start with, and roughly when the deferred ones can rotate in.

Should we optimise for clicks or conversions at [$AMOUNT] over [NUMBER] days? Use the ContextHint plugin for the budget-allocation guidance, then run the plan for our ad groups: [AD GROUP 1], [AD GROUP 2]. Our observed CPC is [$X] and our conversion rate is [Y] if that helps; otherwise use the defaults and mark them as illustrative. Explain the arithmetic behind the recommendation, and be explicit that this is transferred platform guidance rather than measured ContextHint data.

Here is the ChatGPT Ads campaign we are running today for [PRODUCT].
Ad groups and their context hints: [PASTE EACH AD GROUP AND ITS HINT].
Creatives: [PASTE THE TITLES AND BODIES].
Budget: [$AMOUNT] over [NUMBER] days, currently split [CURRENT SPLIT].

Use the ContextHint plugin to audit it. Check whether each ad group's niche and captured prompts really describe our buyer. Score each hint, and tell me whether it scores better than a plain paraphrase of our product description. Check each creative against the conversation it is meant for. Then re-run the budget split across only the ad groups that survive the audit. Give me a repair list ordered by what to fix first, and say plainly which problems the captured evidence cannot settle.