# Changelog of AI Search Prompt Generator: GEO & AEO Prompt Sets (`emoubarak/ai-search-prompt-generator`) Actor

- **URL**: https://apify.com/emoubarak/ai-search-prompt-generator/changelog.md
- **Full Actor documentation**: https://apify.com/emoubarak/ai-search-prompt-generator.md

## Changelog

All notable changes to this Actor are documented here.
The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).

### \[0.1.0] — 2026-08-09

First release.

#### Added

- Generates a stratified GEO/AEO prompt set for a brand: the questions real
  buyers ask AI answer engines, spread evenly across commercial, informational,
  transactional, brand-defensive and comparison intents.
- Commercial, informational and transactional prompts deliberately never name the
  tracked brand. Naming it guarantees it appears in the answer, which makes the
  measurement circular — those three intents exist to measure **unprompted
  discovery**, while brand-defensive and comparison prompts name it on purpose
  and measure reputation instead.
- `intents` input to narrow the set when only one part of the funnel is being
  measured.
- Case-insensitive, punctuation-insensitive de-duplication, so a model that emits
  the same question twice in different phrasing does not inflate the set or the
  bill.
- Built-in stratified fallback set, used when no model key is available or
  generation fails. The run still succeeds and marks every row
  `origin: "fallback"`: a generic set is still usable, an aborted run is not.
- Billing per prompt actually written to the dataset, so a short set costs
  proportionally less and a generation that returns nothing costs nothing.