Demand-led skills for agents that create ad images, product visuals, image edits, and short videos through EvoLink.

The first batch uses scene and intent words as public skill slugs. Seedance 2 and GPT Image 2 are the execution models inside each workflow, not the primary distribution labels.

Seedance 2.5 early access is available here: Get Seedance 2.5 Early Access.

Install the npm package and copy all skills into an agent project:

npx evolink-agent-skills -y --path ./my-agent-project --all

Install one skill at a time:

npx skills add Evolink-AI/evolink-agent-skills --skill image-edit -g
npx skills add Evolink-AI/evolink-agent-skills --skill product-image-generation -g
npx skills add Evolink-AI/evolink-agent-skills --skill image-to-video -g
npx skills add Evolink-AI/evolink-agent-skills --skill product-to-ad -g
npx skills add Evolink-AI/evolink-agent-skills --skill ad-video-generation -g
npx skills add Evolink-AI/evolink-agent-skills --skill product-demo-video -g
npx skills add Evolink-AI/evolink-agent-skills --skill ad-creative-images -g
npx skills add Evolink-AI/evolink-agent-skills --skill ugc-video-ads -g

Each skill also has its own npm package:

npx evolink-image-edit -y --path ./my-agent-project
npx evolink-product-image-generation -y --path ./my-agent-project
npx evolink-image-to-video -y --path ./my-agent-project
npx evolink-product-to-ad -y --path ./my-agent-project
npx evolink-ad-video-generation -y --path ./my-agent-project
npx evolink-product-demo-video -y --path ./my-agent-project
npx evolink-ad-creative-images -y --path ./my-agent-project
npx evolink-ugc-video-ads -y --path ./my-agent-project

All public EvoLink links in this repo use:

utm_source=skillssh&utm_medium=skill&utm_campaign=<skill-name>

Keep campaign names identical to skill slugs so analytics can connect installs, first generation, and token consumption.

1. User installs a task or scene skill from skills.sh.
2. Agent uses GPT Image 2 to create or edit source assets when needed.
3. Agent uses Seedance 2 to generate video from the idea, image, or product context.
4. Agent proposes variants, edits, or platform adaptations.
5. EvoLink attributes token consumption to the skill campaign.

1. image-edit
2. product-image-generation
3. image-to-video
4. product-to-ad
5. ad-video-generation
6. product-demo-video
7. ad-creative-images
8. ugc-video-ads

After each publish, install the skill, verify the skills.sh page, validate all links, and run one real EvoLink generation before releasing the next skill.