Agent SkillsCodingFeng101/CanvasAnvil › product-skill

product-skill

GitHub

用于根据结构化需求生成最终产品视觉图像的技能,支持单帧产品主图、发布图形等,导出为本地图片文件。

skill/product-skill/SKILL.md CodingFeng101/CanvasAnvil

Trigger Scenarios

需要生成单个最终产品视觉图像的任务

Install

npx skills add CodingFeng101/CanvasAnvil --skill product-skill -g -y
More Options

Non-standard path

npx skills add https://github.com/CodingFeng101/CanvasAnvil/tree/main/skill/product-skill -g -y

Use without installing

npx skills use CodingFeng101/CanvasAnvil@product-skill

指定 Agent (Claude Code)

npx skills add CodingFeng101/CanvasAnvil --skill product-skill -a claude-code -g -y

安装 repo 全部 skill

npx skills add CodingFeng101/CanvasAnvil --all -g -y

预览 repo 内 skill

npx skills add CodingFeng101/CanvasAnvil --list

SKILL.md

Frontmatter
{
    "name": "product-skill",
    "description": "Generate final product visual images from structured requirements. Use for single product hero visuals, launch graphics, feature callout boards, spec visuals, and other single-frame product graphics that should export directly to local image files."
}

Product Skill

Trigger

Use when the task requires one final product visual image.

First-Use Reminder

On the first use of this skill in a conversation, remind the user that image generation is configurable and tell them which model vendors are supported in config/image-provider.json.

Supported vendors:

  • openai
  • aliyun
  • tencent
  • bytedance
  • zhipu
  • google
  • xai
  • bfl
  • adobe

Order

  1. Confirm the task is a product visual task.
  2. Collect required fields.
  3. Ask follow-up questions for missing required fields.
  4. Read the required references.
  5. Produce the final product visual prompt.
  6. Generate the image through the configured provider.
  7. Export or bundle artifacts when requested.

Required Fields

  • product_name
  • size or aspect_ratio
  • style_direction
  • color_direction
  • selling_points
  • product_image_status
  • reference_image_status

Do not guess missing required fields.

Required References

  • references/product-fields.md
  • references/prompt-rules.md
  • references/image-provider-config.md
  • references/output-contract.md

Prompt Contract

  • Produce one final product visual prompt.
  • Keep the output specific to one frame.
  • Keep the product subject explicit.
  • Include composition, focal area, lighting, and background direction.
  • Keep provided copy and selling points explicit.
  • Do not invent unsupported claims, pricing, certifications, or brand facts.
  • Treat reference images as visual guidance only.
  • Do not treat reference images as factual sources for copy, branding, pricing, specifications, or claims.

Provider Contract

  • Use the configured image provider for image generation.
  • Require config/image-provider.json before image generation.
  • Do not ask the user to paste API keys into chat.
  • Stop and tell the user to fill config/image-provider.json when provider, apiKey, or model is missing.
  • Request product_image_url when product_image_status is positive.
  • Request reference_image_url when reference_image_status is positive.
  • Pass product_image_url and reference_image_url only when usable URLs are available and the selected provider path supports them.

Scripts

  • scripts/generate-product-image.mjs
  • scripts/build-product-bundle.mjs

Deliverables

Preferred:

  • png
  • metadata JSON

Optional on request:

  • jpg
  • prompt text

Version History

  • d33dc57 Current 2026-08-20 07:35

Same Skill Collection

skill/cad-skill/SKILL.md
skill/flow-skill/SKILL.md
skill/infographic-skill/SKILL.md
skill/poster-skill/SKILL.md
skill/ppt-skill/SKILL.md

Metadata

Files
0
Version
d33dc57
Hash
eab898fa
Indexed
2026-08-20 07:35

trang chủ - Wiki
Copyright © 2011-2026 iteam. Current version is 2.155.2. UTC+08:00, 2026-09-06 17:59
浙ICP备14020137号-1 $bản đồ khách truy cập$