Present generated images, videos, and audio previews in chat for review.
npx skills add https://github.com/Comfy-Org/comfy-cli --skill comfy-relay
Please help me install this Agent Skill. Skill: comfy-relay Source: https://github.com/Comfy-Org/comfy-cli/tree/main/comfy_cli/skills/comfy-relay Command: npx skills add https://github.com/Comfy-Org/comfy-cli --skill comfy-relay
This Skill helps you present creative media work in a way people can actually review, react to, and steer, instead of burying the result in file paths or text descriptions.
Use this skill to show the latest media preview in chat, recommend the strongest version, and keep the next iteration focused on what the user can actually see.
High-intent search queries and answers about installing and using this skill.
To preview generated images directly in chat, the skill presents visual media outputs inline rather than burying them in file paths. This allows you to immediately review renders, compare variants side by side, and guide the next iteration based on visible results.
Reviewing video pacing and structure from comfy-cli workflows is best handled by turning clips into video contact sheets or key frames. This visual inspection method makes it easy to assess pacing, structure, and visual decisions without watching full-length source files.
Yes, you can inspect source blueprints and prompts after seeing the media preview. The skill surfaces the visual preview first for immediate feedback, then shares the prompt or blueprint so you can refine inputs without digging through compiled artifacts.
Async rendering works with audio workflows in comfy-cli by keeping long-running jobs responsive. The skill supports image, video, and audio workflows, ensuring that outputs remain accessible in the project directory while you continue other tasks.
Perceptual limits and artifacts are handled through honest reporting and careful artifact management. The skill ensures outputs remain accessible in the project directory and transparently communicates any perceptual limitations rather than masking visual discrepancies.
Alexander-TyagunovInitialize Magician workspaces with directory structure, CLAUDE.md, and permissions allowlist.
shekerkammaGenerate multiple UI design variants from a creative brief.
Smooth-OperationDispatch independent tasks to parallel agents and track progress.
hidonguyenAnalyze and generate multimodal media using Gemini and MiniMax APIs.
