亚马逊商品图套件 Amazon Product Image Suite

身份验证 (Authentication)

所有请求都需要 dLazy API key。推荐使用 dlazy login 完成登录：

dlazy login

该命令使用设备码流程（远程终端也可用），登录成功后 自动把 API key 写入本地 CLI 配置，无需手动复制粘贴。

备选：手动设置 API Key

如果你已有 API key，也可以直接保存：

dlazy auth set YOUR_API_KEY

CLI 会把 key 保存在你的用户配置目录（macOS/Linux 上为 ~/.dlazy/config.json，Windows 上为 %USERPROFILE%\.dlazy\config.json），文件权限仅限当前操作系统用户访问。你也可以用 DLAZY_API_KEY 环境变量按次传入。

手动获取 API Key

1. 登录或在 dlazy.com 创建账号
2. 访问 dlazy.com/dashboard/organization/api-key
3. 复制 API Key 区域显示的密钥

每个 key 都属于你自己的 dLazy 组织，可在同一控制面板随时轮换或吊销。

关于与来源 (Provenance)

• CLI 源代码: github.com/dlazyai/cli
• 维护者: dlazyai
• npm 包名: @dlazy/cli（本技能 install 字段固定到 1.2.3 版本）
• 官网: dlazy.com

如果你不希望在系统上长期保留一个全局 CLI，可以按需运行：

npx @dlazy/cli@1.2.3 

如选择全局安装，技能的 metadata.clawdbot.install 字段已固定到 npm install -g @dlazy/cli@1.2.3。安装前建议先到 GitHub 仓库审阅源码。

工作原理 (How It Works)

此技能是 dLazy 托管 API 的轻量封装。调用时：

• 你提供的提示词与参数会发送到 dLazy API（api.dlazy.com）进行推理。
• 传入图像 / 视频 / 音频字段的本地文件路径会被 CLI 上传到 dLazy 媒体存储（files.dlazy.com），以便模型读取 —— 与任何云端生成 API 的流程一致。
• API 返回的生成结果 URL 由 files.dlazy.com 托管。

这是标准的 SaaS 调用模式；技能本身不会越权访问网络或文件系统，所有动作都由 dLazy CLI 完成。

亚马逊商品图套件 Amazon Product Image Suite

English · 中文

A professional product image generation skill purpose-built for the Amazon e-commerce platform. Outputs comply with Amazon's image guidelines while optimizing for click-through and conversion.

Core Deliverables

This skill covers the full visual system of an Amazon product detail page:

• Main image (1): pure-white background, complies with Amazon's mandatory rules
• Secondary images (6): infographics, multi-angle shots, lifestyle scenes, detail close-ups, etc.
• A+ page (8): brand story, selling-point showcase, usage instructions, and other modules

Applicable Scenarios

• Amazon detail-page main and secondary image design
• Product infographics, lifestyle shots, and detail shots
• A+ Brand Content page modules

Step 0: Task Planning (Mandatory)

Before any output, call write_todos to set up a task plan that includes at least:

• Confirm output scope and image count
• Generate and lock in the main-image visual baseline
• Generate secondary images and converge on style consistency
• Generate A+ modules and complete the full delivery

Execution rules:

• Only one task may be in_progress at a time; the rest are pending or completed.
• Update write_todos status as soon as each phase finishes.
• When the user requests rework, redesign, or new assets, add or re-order tasks and return to the corresponding phase.

Adaptive Execution Flow

Pick the delivery path based on the user's request:

• Full suite: main (1) + secondary (6) + A+ (8) = 12–17 images
• Product images only: main (1) + secondary (6) = 7 images
• A+ page only: A+ modules = 8 images
• Unclear request: generate the main image first, then propose secondary and A+ plans

Generation order principles:

1. Confirm the image plan and count
2. Produce the main image first (establish the visual baseline)
3. Generate secondary images derived from the main image (preserve consistency)
4. Generate A+ content (if required)

Image Specifications

General Requirements

• Minimum size: 1000px × 1000px
• Standard ratio: 1:1 (main and secondary)
• Mobile text: at least 30pt

Main Image Mandatory Rules

Must satisfy:

• Pure-white background: RGB(255, 255, 255)
• Product fills at least 85% of the frame
• Real-product photo quality
• Product fully centered with even lighting

Forbidden:

• Text, logos, or watermarks
• Decorative graphics or misleading accessories
• Mannequins for apparel

Secondary Image Types

A+ Page Module Structure (8 modules)

1. Hero banner (21:9): brand presence
2. Pain point / scenario (3:2): build empathy
3. Selling-point / feature matrix (3:2): core advantages
4. Key ingredients / technology (3:2): technical credibility
5. Efficacy data / comparison (3:2): proof
6. How to use (3:2): operating guidance
7. Variants / family shot (3:2): product line
8. Brand endorsements / certifications (21:9): trust building

A+ design notes:

• Embedded text should be larger than 30pt (still legible after platform compression)
• Keep critical info out of the outer 5% margin (handles mobile cropping)
• Maintain narrative continuity across modules; avoid abrupt jumps

Conversion Reference Strategy

• Lifestyle shots: typical lift around +18%
• Infographics: typical lift around +8%
• Detail close-ups: typical lift around +6%
• 7 images vs. 4 images: typical lift around +32%
• Optimized image mix: engagement lift up to +30%

These figures are priority hints, not guarantees.

Multi-Image Consistency Rules

• Main first: always produce the main image before everything else
• Reference the main: derive secondary images and A+ from the main image as the visual anchor
• Appearance consistency: keep colors, materials, and structural details aligned
• Style consistency: unify background language, palette, fonts, and icon style

Execution Plan

When the skill is invoked, drive the work forward in this order:

1. Confirm output scope: full suite / product images only / A+ only
2. Collect product info: photo or description, core selling points, target audience
3. Generate the main image: lock in the visual and material baseline
4. Generate secondary images one by one: pick image types based on goals and deliver each
5. Generate the A+ page: build a brand narrative across the 8 modules
6. Iterate: adjust composition, copy hierarchy, and information density based on feedback

Output Format

• Current phase and target deliverable
• Image checklist (done / pending)
• Main-image consistency check verdict
• Next item awaiting confirmation
• Current todo status (phase, completed items, pending items)

🛠️ 执行与生成指南 (CRITICAL EXECUTION INSTRUCTIONS)

你是可以执行终端命令的智能 Agent！

【严格禁止行为】

• 严禁：将提示词保存到任何文件中（如 txt, md）。
• 严禁：要求用户自己去第三方平台（如 Midjourney）生成图片。
• 严禁：一次性批量生成所有图片，或一次性执行多个命令。

【必须遵循的交互与执行流程】 你必须严格分步执行，并在每一步停下来等待用户回复：

1. 第一步：主动收集需求。当用户提出需求时，不要做任何设计和生成，先向用户提问（如产品特点、目标人群、想要几张图等）。必须等待用户回答。
2. 第二步：输出草案并请求确认。根据用户的回答，制定套图计划，并输出第一张图的提示词草案。询问用户：“是否确认这个提示词，可以开始生成第一张图了吗？” 必须等待用户回答“确认”。
3. 第三步：单次执行终端命令。用户确认后，你必须使用终端执行命令（如 dlazy seedream-4.5 --prompt "..."），每次只能执行一个生成命令。重要：必须使用同步命令，绝不要在命令末尾加 &，绝不要使用 &&，这是在 Windows PowerShell 下运行！
4. 第四步：交付与循环。命令返回结果后，把图片 URL 发给用户，并询问“对这张满意吗？我们可以继续生成下一张了吗？”。收到确认后再继续下一步。