分镜图工作流 Image Storyboard

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/[email protected] 

如选择全局安装，技能的 metadata.clawdbot.install 字段已固定到 npm install -g @dlazy/[email protected]。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。

---

name: 'image-storyboard' description: 'A professional storyboard skill that turns user creative briefs into industry-standard storyboards, covering film, advertising, short video, and educational narrative scenarios.'

分镜图工作流 Image Storyboard

English · 中文

A professional storyboard skill for film, advertising, short video, and educational narrative scenarios, built around a strict "plan first, render later" flow.

Core Positioning

Turn user creative briefs into industry-grade storyboards across two main pipelines:

• Cinematic storyboards: film, advertising, shorts
• Narrative storyboards: educational content, comic-style narrative

Step 0: Task Planning (Mandatory)

Before any execution, set up a task plan that includes at least:

• Requirements exploration and tech-spec lock-in
• Character design and character master sheet confirmation
• Script structuring and script-gate confirmation
• Image generation and batched delivery
• Final storyboard assembly and export

Execution rules:

• Only one task may be in_progress at a time; the rest are pending or completed.
• Update the task plan as soon as each state finishes.
• When the user asks to roll back or rework, add or re-order tasks and return to the corresponding state.

Technical Specification System

1) Cinematic Storyboards (Film / Ad / Short)

• Aspect ratios: 16:9, 2.35:1, 9:16, etc.
• Required metadata:
  • Shot size (close-up, medium, wide, etc.)
  • Camera movement (push, pull, pan, dolly, tracking)
  • Lighting and color temperature

2) Narrative Storyboards (Education / Comic)

• Aspect ratios: comic or vertical-narrative standards
• Required metadata:
  • Sequence markers (e.g., S01-P03)
  • Mood tags (tense, warm, relieved, etc.)

3) Character Master Sheet

• Visual standard: clean full-body reference, no text, frames, or UI elements
• Style fit:
  • Cinematic projects prefer photoreal style
  • Narrative projects prefer 2D or sketch styles
• Core elements:
  • Identity, age, vibe
  • Appearance, costume details, key accessories
• Generation formula:
  • [subject] + [character traits] + [tech specs] + [style] --no text

4) Script Structure Format

Every storyboard panel must output in this structure:

• Visual prompt: [subject/character] + [action/interaction] + [environment/scene]
• Tech params: [shot size, camera, lighting, etc.]
• Text/logic: [narrative or logic elements]
• Status: [pending / generated]

5) Delivery Spec (final storyboard)

• Paper size: A4 landscape (297mm × 210mm)
• Output: print and PDF export
• L1 layout (vertical film strip, cinematic): black background, single column
• L2 layout (comic grid, narrative): white background, thick borders, caption boxes, step markers

Standard Workflow (4 states)

Every reply must start with:

• **Current Step:** [state] | **Next:** [goal]

State 1: Requirements Exploration and Validation

Goal: lock in tech specs and the style guide.

Execution requirements:

1. Analyze the user's input and fill in missing info.
2. Determine aspect ratio, panel count, and category (cinematic or narrative).
3. Output as a bulleted list, not a table.
4. Wait for the user's explicit "continue" or "confirm."

Suggested prompt:

• The specs are organized — confirm to proceed to the character design phase?

State 2: Character Design (Visual Bible)

Goal: lock in character appearance to prevent later drift.

Execution requirements:

1. Build the character visual bible: appearance, costume, accessories, posture baseline.
2. Generate and present a single character master reference image.
3. Output as a bulleted list.
4. Strict gate: you must wait for character approval before entering the script phase.

Suggested prompt:

• Character look is locked — start the storyboard script?

State 3: Storyboard Production (Script and Visuals)

Phase 1: Script Writing (think)

• Convert the narrative into a structured script.
• Output as a bulleted list.
• Strict gate: do not generate images until the script is approved.

Output Requirements

• Use bulleted lists, not tables (except for the delivery layout spec).
• Clearly mark current state, completed items, and items awaiting confirmation.
• All critical gate nodes must wait for user confirmation before continuing.

🛠️ 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。