gen-ai-skills
Shared skills, prompts, and tooling for working with Picsart's gen-ai CLI and related generative-AI workflows.
This repository is a public, community-friendly home for reusable skills and recipes. It is not the source of the gen-ai CLI itself — it complements it.
Skills here are consumable by any agent that reads SKILL.md files: Claude Code, Cursor, Codex, and others.
Install
npx skills add PicsArt/gen-ai-skills
See INSTALL.md for gh skill install, the Claude Code marketplace, and the ./setup script alternatives.
Repository layout
gen-ai-skills/
├── .claude-plugin/ # Claude Code plugin + marketplace manifests
├── .codex-plugin/ # Codex plugin manifest
├── .cursor-plugin/ # Cursor plugin manifest
├── .github/workflows/ # CI (security scan, lint)
├── scripts/ # Repo tooling (scan-skill.py, ...)
├── skills/ # All skills live here, one directory each
│ ├── gen-ai-use/
│ │ └── SKILL.md
│ ├── marketer-campaign-kit/
│ │ └── SKILL.md
│ └── ...
├── INSTALL.md
├── CONTRIBUTING.md
├── VERSION
├── setup # universal symlink installer
└── README.md
Each skill lives in its own directory under skills/. The directory name is the install name. The directory must contain a SKILL.md with YAML frontmatter (name, description, optional argument-hint, optional allowed-tools) followed by the skill body.
Skills
This bundle ships 20 skills. Several skills are mode-routers — one entry point with references/modes/ for variants — to keep the catalog small and trigger-friendly.
Core CLI
| Skill |
Description |
gen-ai-use |
Generate AI images, videos, audio via the Picsart gen-ai CLI. |
gen-ai-persona-creation |
Create AI influencers, branded characters, or pet personas from a brief or reference. |
motion-studio |
End-to-end AI video production: references, clips, scene assembly, audio. |
API
| Skill |
Description |
picsart-api |
Call Picsart's image, video, GenAI, and variable-data REST APIs via the picsart MCP server. |
Multi-mode skills
| Skill |
Description |
product-photo-studio |
Transform product photos. Modes: bulk-restyle, compose, seasonal, variants, reshoot, mockups. |
text-to-visual |
Generate visuals from text. Modes: single image, article-set, OG image. |
multi-channel-bundle |
Ship a coordinated multi-format bundle. Modes: campaign, launch. |
Agency
Dev
Enterprise
Marketer
Prosumer
More skills are welcome — see CONTRIBUTING.md.
License
MIT
No comments yet
Be the first to share your take.