Open Image Prompts skill
A local archive of image prompts with two skills: an art-direction brief and prompt search with a sample gallery.
- 425 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The skill downloads a database and images from GitHub Releases and starts a local gallery on 127.0.0.1.
- The database consists of other people's prompts and images: third-party rights are described in `DATA_LICENSE.md`.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/open-image-promptsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The skill downloads a database and images from GitHub Releases and starts a local gallery on 127.0.0.1.
- The database consists of other people's prompts and images: third-party rights are described in `DATA_LICENSE.md`.
This is third-party code. Review the repository files before installing.
What it does
Open Image Prompts gives an agent two skills. `img-gen-taste` turns a rough brief into a clear art direction and works at once, from its bundled style cards. `img-gen-prompts` searches past prompts together with their images in a local database and opens a gallery for comparing samples. The database holds more than 22 thousand prompts, about 38 thousand images and translations, labeled with a closed set of 185 visual labels. It is stored in SQLite, downloaded from GitHub Releases and opened read-only. Search works on the meaning of the request and declared preferences, with no vector database, model download or API key.
Who it is for. Designers and marketers who pick a prompt for image generation and want to lean on past successful samples.
Good fit when
- You have a rough brief such as «a vintage city travel poster» and need to turn it into a style direction.
- Before generating an image you want to see sample prompts with real results next to them.
- References for a series of images are picked and compared in a gallery.
- Search runs against a local database, with no internet and no API key.
Not a fit when
- Image generation itself: the skill picks prompts and references, while another model draws the image.
- Little disk space: the full image set weighs about 4.3 GB and the database about 80 MB.
- Offline video viewing without a separate download: the packs hold posters, not the clips themselves.
Example request
Pick a style direction for a vintage travel poster and find five similar prompts with images.Limitations
Search needs the downloaded database: a clone without it holds only code. `npm run data:pull` fetches the database (about 80 MB) and image packs (about 4.3 GB), while `npm run data:pull:db` fetches only the database and the gallery then loads images from their original links. Git, Node.js 20.19 or newer (or 22.12 or newer) and Python through uv are required. The packs hold no clips and the gallery shows posters. For playback `scripts/fetch_videos.py` downloads the videos separately, about 8.8 GB.
How to disable. Delete the `img-gen-taste` and `img-gen-prompts` folders from the agent's skills directory. Downloaded data lives in `db/`, `images/` and `.oip/` inside the clone and goes away with it.
FAQ
How do I install the skills?
With `npx skills add NanmiCoder/open-image-prompts -g`. `npx skills add NanmiCoder/open-image-prompts --list` shows the list without installing.
Do I need to download the database?
Yes, for `img-gen-prompts`. `npm run data:pull:db` fetches only the database, and `npm run data:pull` adds the images. `img-gen-taste` works at once.
How do I check that everything is ready?
Point `OIP_REPO_ROOT` at the clone and run `npm run status`. A ready install prints `"ready": true` and the `oip-visual-v2` taxonomy.
Do I need an API key or a model?
No. Search runs on a local SQLite database with no vector database, model download or API key.
What license covers the data?
Code and skill instructions are MIT. The data has separate terms, described in the repository's `DATA_LICENSE.md`.
Discussion
No questions or reviews yet.
Related
SkillsA skill that searches a library of 10,000+ image prompts, shows sample images and tailors the chosen prompt to your article or script
A library of 150 GPT Image prompts, each credited to its author with a link to the post, grouped into nine use cases
Awesome GPT Image
awesome-gpt-image
A collection of the best prompts and examples for GPT Image 2 and 2.5 gathered from well-known authors on X: ready request texts for images in different styles
A skill that turns a picture into a generator prompt: frame breakdown, an enhanced version, three variants and platform editions.
A Claude Code skill that compiles a vague request into a finished gpt-image-2 prompt and checks it with a validator script
GPT Image 2 Prompts by YouMind
awesome-gpt-image-2
A prompt library for OpenAI GPT Image 2: more than 2000 curated prompts with previews, updated daily by the author's account
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| Open Image Prompts (this card) | no | 425 | |
| GPT-Image2 Style Library | no | 35,752 | picks a template from GPT-Image2 cases and assembles a prompt, while Open Image Prompts searches a local database of labeled images |
| AI Image Prompts | no | 1,377 | also searches a prompt library, while Open Image Prompts keeps the whole database on your disk and labels images with a closed taxonomy |
README in short
The repository tracks the code, interface, skills and taxonomy, and ships the heavy database and image files as separate releases with sha256 checks. A condensed AGENTS.md covers setup, ports and the skill contract for agents.