MeiGen launches free prompt gallery with Claude MCP and Figma hooks

• hasantoxr's launch post framed MeiGen as a free prompt gallery that pulls together viral prompts for GPT Image 2, Nano Banana 2, Seedance 2.0, Veo 3.1, and Midjourney.
• According to hasantoxr's feature list, MeiGen adds drag-to-canvas prompt application, reverse prompting from a reference image, model and category filters, and daily free credits.
• hasantoxr's workflow post said the project already reaches beyond a gallery, with an MCP server for Claude, a Figma plugin, and an OpenClaw skill.
• hasantoxr's open-source post said the dataset is fully open source, while the linked GitHub repo suggests the team wants other people to fork or build on top of it.

You can jump straight to the live site, inspect the GitHub repo, and watch the gallery walkthrough cycle through clay figurines, museum-style infographics, fashion ads, and worldbuilding prompts. The useful detail in hasantoxr's product tour is the workflow angle: prompt cards can be applied directly on a canvas, and a reference image can be turned back into a prompt. Then hasantoxr's integration post pushes it into tooling, with Claude, Figma, and OpenClaw already in the thread.

The pitch is simple: MeiGen is indexing prompts that already performed on X, then surfacing them as a searchable gallery instead of a paid prompt pack.

According to hasantoxr's category list, the early catalog includes 3D figurine and clay looks, cinematic editorial campaigns, Chinese museum-style infographics, red carpet fashion ads, Y2K cosmetic posters, and larger worldbuilding sets.

That makes the product feel closer to a creative reference board than a plain prompt dump. The source material is organized around visual outcomes people already shared, not around abstract prompt-writing theory.

The most concrete product mechanics are all in hasantoxr's feature list, which breaks MeiGen into four actions:

• Drag a prompt card onto the canvas to apply it instantly.
• Drop in a reference image and have the app reverse-engineer a prompt.
• Filter by model, category, and what is trending that day.
• Use daily free credits instead of a paid unlock.

The reverse-prompting piece is the interesting one for designers. It turns the gallery into a two-way tool, part archive, part prompt extractor.

The thread claims MeiGen is already wired into three other surfaces:

• An MCP server that plugs into Claude.
• A Figma plugin for design workflows.
• An OpenClaw skill for agent-driven generation.

If those hooks hold up, the gallery is less about browsing and more about reusing prompt patterns inside the tools people already work from.

The last useful reveal is distribution. hasantoxr's open-source post said the full dataset is open source, and linked directly to a public GitHub repository.

That matters because the thread is not selling exclusivity. It is explicitly positioning the prompt set as something people can copy, fork, or turn into their own products, with the original creators credited and view counts visible, as hasantoxr's follow-up post put it.