← 返回 Skills 市场
WeryAI Image Generator
作者
WeryAI-Developer
· GitHub ↗
· v0.1.5
· MIT-0
162
总下载
0
收藏
0
当前安装
6
版本数
在 OpenClaw 中安装
/install weryai-image-generator
功能描述
Generate WeryAI images from text prompts or reference images through the WeryAI image APIs. Use when the user needs text-to-image, image-to-image, async imag...
安全使用建议
This package appears to be a legitimate WeryAI client. Before installing or running it: 1) Treat your WERYAI_API_KEY like any secret — provide it only to trusted environments and never commit it to source. 2) Be careful with local-file inputs: if the agent receives or is asked to use a filesystem path, the script will read that file and upload it to the configured WeryAI host. Do not let the agent guess paths or accept unvetted user-provided paths that might point at sensitive files. 3) Double-check the endpoint env vars: the code defaults to api.weryai.com but metadata lists WERYAI_BASE_URL/WERYAI_MODELS_BASE_URL as required and you can override them; only set them to trusted domains. Avoid setting WERYAI_ALLOW_INSECURE_UPLOAD (or any flag that suppresses warnings) unless you fully trust the target host. 4) If you need higher assurance, review the scripts (especially uploadLocalFileToPublicUrl and createClient) or run the dry-run commands first (models-image.js and wait-image.js --dry-run) to confirm behavior without spending credits.
功能分析
Type: OpenClaw Skill
Name: weryai-image-generator
Version: 0.1.5
The skill bundle is a legitimate and well-structured integration for the WeryAI image generation service, supporting text-to-image, image-to-image, and task status monitoring. It includes robust features such as local file uploading (with extension whitelisting and size limits), dynamic model capability discovery from official endpoints (api.weryai.com), and bounded polling logic to prevent infinite execution loops. The code follows standard security practices for agent skills, such as using environment variables for sensitive API keys and providing dry-run modes for payload verification, with no evidence of malicious intent, data exfiltration, or prompt injection.
能力评估
Purpose & Capability
Name/description, required binaries (node), and required env vars (WERYAI_API_KEY, WERYAI_BASE_URL, WERYAI_MODELS_BASE_URL) align with an API-backed image-generation client. The provided scripts implement text-to-image, image-to-image, status polling, model lookup and balance checks — all expected for this purpose.
Instruction Scope
SKILL.md and the scripts focus on submitting generation tasks, bounded polling, model inspection and returning images. The runtime supports local file paths which will be automatically uploaded to the WeryAI service (uploadLocalFileToPublicUrl). That behavior is documented in SKILL.md, but it means any local path the agent is given can be read and uploaded — a potential data-exfiltration vector if a user (or an automated decision) supplies sensitive filesystem paths.
Install Mechanism
There is no external install/download step; this is an instruction+script bundle that requires Node on PATH. No remote archive downloads or third-party install hooks were found in the package metadata.
Credentials
Most required env vars are proportional (API key and API endpoints). However: (1) SKILL.md states WERYAI_BASE_URL and WERYAI_MODELS_BASE_URL have defaults but the metadata declares them as required — that is inconsistent; (2) the code references an additional env var WERYAI_ALLOW_INSECURE_UPLOAD (used to suppress a warning when uploading to non-official domains) which is not declared in requires.env. Together these allow the runtime to be pointed at arbitrary hosts (and to silence the warning), which increases risk if misconfigured.
Persistence & Privilege
The skill is not marked always:true and does not request system-wide config changes. It will run as a normal user-mode tool and only acts when invoked. It does read local files only when given explicit local paths.
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 weryai-image-generator - 安装完成后，直接呼叫该 Skill 的名称或使用
/weryai-image-generator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
v0.1.5
weryai-image-generator v0.1.5
- Added support for `WERYAI_BASE_URL` and `WERYAI_MODELS_BASE_URL` environment variables in required setup.
- Updated documentation: The skill now explicitly supports local file path inputs for reference images, with automatic uploading to WeryAI servers.
- Security section improved to clarify that local file paths are handled safely.
- No code changes detected; all updates are documentation and metadata clarifications.
v0.1.4
Initial release with full code support for WeryAI image generation.
- Added all code and scripts for text-to-image and image-to-image workflows using WeryAI APIs.
- Supports image task submission, polling for output, task status checks, and one-shot wait behavior if requested.
- Allows model switching, dry-run previews, and capability queries.
- Handles local and URL-based reference images with auto-upload.
- Provides model capability lookup, default parameters, and user-overridable settings.
v0.1.3
weryai-image-generator v0.1.3
- Updates user guidance: When asked "how to use this skill", respond with natural language and prompt examples instead of pasting raw shell commands.
- Minor wording improvements in the description and guidance sections for clarity.
- No code changes or new features.
v0.1.2
**Major restructuring and simplification.**
- Removed 20 internal scripts related to image generation tasks, model and status checks.
- Updated documentation to clarify that the skill now uses an image-first workflow with bounded status polling (max 5 minutes).
- Reduced emphasis on returning task/batch IDs; images are the default output unless user requests otherwise.
- Documentation now aligns with new async/polling flow and simpler agent integration.
- Retains authentication, quick verification, and parameter guidance instructions.
v0.1.1
- Switches default workflow to asynchronous: submitting image generation tasks now returns a task/batch ID, with separate status polling. One-shot wait mode only when explicitly requested.
- Supports agent environments by decoupling submission and result polling through new script structure (`submit`, `status-image.js`, `wait-image.js`).
- Adds local image file upload for image-to-image: automatically uploads non-HTTP image sources.
- Removes old synchronous scripts and internal model/setup tools, adds new streamlined script files for async flow and shared logic.
- Documentation updated to reflect async pattern, input media handling, and improved guidance for user parameter mapping.
v0.1.0
WeryAI Image Generator skill v0.1.0 – initial release
- Supports text-to-image, image-to-image, restyle, reference-image generation, and model/task management using WeryAI APIs.
- Defaults to WERYAI_IMAGE_2_0 with 1 image at 9:16 aspect ratio; users can override model and parameters.
- Includes dry-run payload previews, model capability lookup, and task status polling for safe and flexible operation.
- Requires WERYAI_API_KEY and Node.js >=18; guides user through setup and image/request validation.
- Offers operator-style guidance, with default settings mapped intelligently from user requests and safe confirmation before paid runs.
元数据
常见问题
WeryAI Image Generator 是什么？
Generate WeryAI images from text prompts or reference images through the WeryAI image APIs. Use when the user needs text-to-image, image-to-image, async imag... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 162 次。
如何安装 WeryAI Image Generator？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install weryai-image-generator」即可一键安装，无需额外配置。
WeryAI Image Generator 是免费的吗？
是的，WeryAI Image Generator 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
WeryAI Image Generator 支持哪些平台？
WeryAI Image Generator 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 WeryAI Image Generator？
由 WeryAI-Developer（@weryai-developer）开发并维护，当前版本 v0.1.5。
推荐 Skills