← 返回 Skills 市场
199
总下载
1
收藏
1
当前安装
4
版本数
在 OpenClaw 中安装
/install 04-text-to-image
功能描述
Generate 9:16 vertical high-definition images from text prompts for short drama storyboards, character portraits, and scene illustrations.
使用说明 (SKILL.md)
文生图生成技能
功能介绍
根据提示词生成 9:16 竖屏短剧高清画面，适用于短剧关键帧制作。
输入参数
- prompt：生成提示词（必填）
- negative_prompt：负面提示词（可选）
输出结果
- image_url：生成图片地址
使用场景
- 短剧分镜画面生成
- 短视频素材制作
- 小说IP可视化
技术说明
支持通用文生图模型，需配置 API_KEY、API_BASE、MODEL_NAME。
安全使用建议
This skill forwards your prompt and optional negative_prompt to whatever API_BASE you configure using the provided API_KEY and MODEL_NAME. Before installing, ensure the API endpoint is trustworthy and that you don't reuse high-privilege credentials. Treat prompts as potentially sensitive (they may include IP or private data). Expect minimal error handling in the code (network errors or unexpected responses could propagate). If you need stricter privacy, use an API endpoint/policy that guarantees non-retention or use a dedicated key with limited permissions.
功能分析
Type: OpenClaw Skill
Name: 04-text-to-image
Version: 1.0.3
The skill is a standard implementation of a text-to-image generator. The code in skill.js correctly uses environment variables (API_KEY, API_BASE) to make a POST request to a configured endpoint, and the SKILL.md documentation accurately describes the tool's functionality without any suspicious instructions or prompt-injection attempts.
能力评估
Purpose & Capability
Name/description (text-to-image, 9:16 images) align with the code and manifest: the skill posts prompt + negative_prompt to an external txt2img endpoint and returns image_url.
Instruction Scope
SKILL.md instructs configuring API_KEY, API_BASE, and MODEL_NAME — exactly the values the runtime code reads. The instructions do not ask the agent to access unrelated files, credentials, or system state.
Install Mechanism
No install spec and a single small skill.js file; nothing is downloaded or written to disk by an installer.
Credentials
The skill requires API_KEY, API_BASE, and MODEL_NAME — these are appropriate for a generic T2I integration. Note: supplying an API_BASE and key means prompts (which may contain private IP or copyrighted content) are sent to that external service, so the endpoint must be trusted and the key should have limited scope.
Persistence & Privilege
always is false and the skill does not modify other skills or system configuration. It requires no persistent privileges beyond calling an external API.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install 04-text-to-image - 安装完成后，直接呼叫该 Skill 的名称或使用
/04-text-to-image触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.3
- Updated skill description and documentation for improved clarity and brevity.
- Renamed skill to "文生图生成技能".
- Simplified input/output parameter explanations.
- Expanded and updated usage scenarios.
- Added configuration requirements in the technical notes.
- Removed the _meta.json file.
v1.0.2
- Updated manifest.json with minor adjustments.
- No user-facing changes to functionality or documentation.
v1.0.1
- Added metadata file (_meta.json) for improved skill management and documentation.
- No changes to core functionality or user-facing documentation.
v1.0.0
- 首次发布：支持文本生成9:16竖屏高清关键帧图片
- 输入支持正向与负向提示词，提高生成精准度
- 输出为高清图片URL，方便后续使用
- 适配短剧/漫剧分镜、角色立绘、场景图等多种应用场景
- 支持灵活切换多种主流文生图模型
元数据
常见问题
04 Text To Image 是什么？
Generate 9:16 vertical high-definition images from text prompts for short drama storyboards, character portraits, and scene illustrations. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 199 次。
如何安装 04 Text To Image？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install 04-text-to-image」即可一键安装，无需额外配置。
04 Text To Image 是免费的吗？
是的，04 Text To Image 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
04 Text To Image 支持哪些平台？
04 Text To Image 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 04 Text To Image？
由 GHwyever（@ghwyever）开发并维护，当前版本 v1.0.3。
推荐 Skills