Version Log
What changed, when, and what's being considered next. Current version: v1.3.0.
v1.3.0 — 2026-08-09
- Added a first-visit popup pointing new visitors to the Disclaimer and How to Use pages.
- Launched the public Q&A wall — post a question, feedback, or whatever's on your mind.
- Added this Version Log page.
v1.2.0 — 2026-08-09
- Added the How to Use guide, covering every field on the Builder screen station by station.
- Added the Disclaimer page clarifying this is a prompt builder, not a generator — with a link out to Magnific for anyone who wants to skip straight to generating an image.
- Fixed a site-wide mobile bug where the header overflowed and forced horizontal scrolling on phones.
v1.1.0 — 2026-08-09
- Text-mode prompts (ChatGPT and similar) now assemble as real instructions — Persona, Tone, Output Format, and Reasoning Style become actual sentences instead of comma-separated tags.
- Selected style terms now follow each category's defined order (Medium → Camera → Lighting → ...) instead of whatever order they were clicked in.
- Duplicate ideas are automatically skipped if they already appear in the Main Idea or an earlier part of the prompt.
- Fixed de-emphasized terms reading as "light X" (sounded like a lighting instruction) — now "faint suggestion of X".
v1.0.0 — 2026-08-08
- Initial launch: multi-provider prompt Builder (image, video, audio, text), the Library, the Admin panel, and the Buy Me a Coffee page.
Up Next / Under Consideration
Not promises, not dated — just what's on the radar.
- Flagging conflicting terms picked in the same category (e.g. "daytime" and "nighttime" both selected).
- Provider-aware length/token limits, so an overly long prompt gets flagged before you hit a model's cap.