← Back to Skills Marketplace
1237
Downloads
3
Stars
4
Active Installs
1
Versions
Install in OpenClaw
/install qwen-image-skill
Description
Generate images using Alibaba Cloud Bailian Qwen-Image and Z-Image models (通义千图文生图 + 人像照片模型)
Usage Guidance
This skill appears to implement the described Alibaba Bailian image-generation flow and only needs a DASHSCOPE_API_KEY. Before installing: 1) verify the package metadata vs SKILL.md — ensure you supply DASHSCOPE_API_KEY if you intend to use it; 2) prefer setting the API key as an environment variable (export DASHSCOPE_API_KEY) rather than relying on TOOLS.md, since the script will try to read TOOLS.md from several filesystem locations (including unexpected paths) which could expose local files; 3) consider creating a dedicated API key with limited scope/quota for this skill; 4) be aware the script makes outbound HTTPS requests to dashscope.aliyuncs.com and will download the returned image from whatever host the API provides; 5) if you are uncomfortable with the script reading files outside the skill directory, inspect or run it in an isolated environment (container or VM) or modify the script to restrict TOOLS.md lookup to a single safe location. If these issues are acceptable, the skill is plausibly coherent with its stated purpose; otherwise proceed with caution.
Capability Analysis
Type: OpenClaw Skill
Name: qwen-image-skill
Version: 1.0.0
The skill is designed to generate images using Alibaba Cloud's Qwen-Image and Z-Image models. The `SKILL.md` provides clear instructions and usage examples without any prompt injection attempts. The `scripts/generate.py` script retrieves the `DASHSCOPE_API_KEY` from environment variables or the `TOOLS.md` file (a documented configuration source for OpenClaw skills), makes HTTPS requests to the legitimate Alibaba Cloud API endpoint `dashscope.aliyuncs.com`, and saves the generated image locally. All observed behaviors are aligned with the stated purpose, and there is no evidence of data exfiltration to unauthorized endpoints, malicious execution, persistence mechanisms, or obfuscation.
Capability Assessment
Purpose & Capability
The skill's purpose (generate images via Alibaba Bailian Qwen-Image / Z-Image) matches the code and API endpoint used (dashscope.aliyuncs.com). However, the registry metadata in the package summary lists no required environment variables while SKILL.md and the script declare and require DASHSCOPE_API_KEY — an inconsistency that should be resolved.
Instruction Scope
Runtime instructions and the included script are focused on image generation and are proportionate: they build a JSON payload, POST to the Bailian endpoint, parse the response, and download the returned image. The script also attempts to discover the API key by reading TOOLS.md from several filesystem locations (current working dir, an absolute path /home/admin/clawd/TOOLS.md, and a path four levels up from the script). Reading those files is intended to obtain the declared key, but the locations accessed could expose or read files outside the skill's folder — worth noting before granting access.
Install Mechanism
This is an instruction-only skill with a single Python script and no install spec or external downloads. There is no package install or archive extraction — low install-surface risk.
Credentials
The only credential the skill needs is DASHSCOPE_API_KEY (used as a Bearer token to call the Bailian API), which is reasonable for this purpose. But the mismatch between the package-level 'required env vars: none' and SKILL.md's 'DASHSCOPE_API_KEY' is confusing. Also, the script's fallback to searching multiple TOOLS.md locations may read user files containing other secrets if present — it only extracts a matching DASHSCOPE_API_KEY line, but file reads are sensitive.
Persistence & Privilege
The skill does not request always:true, does not modify other skills or system config, and does not install persistent components. It only performs network calls at runtime and writes the downloaded image to a local path specified by the user (or a generated filename).
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install qwen-image-skill - After installation, invoke the skill by name or use
/qwen-image-skill - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Qwen-Image skill (v1.0.0) initial release:
- Generate images using Alibaba Cloud's Qwen-Image and Z-Image models.
- Automatic scene detection chooses the best model (e.g., z-image-turbo for portraits, qwen-image-max for general use).
- Supports high-quality Chinese text rendering in images.
- Multiple aspect ratios and model options available, with manual override support.
- Simple setup—requires only a DASHSCOPE_API_KEY.
Metadata
Frequently Asked Questions
What is qwenz-image-gen?
Generate images using Alibaba Cloud Bailian Qwen-Image and Z-Image models (通义千图文生图 + 人像照片模型). It is an AI Agent Skill for Claude Code / OpenClaw, with 1237 downloads so far.
How do I install qwenz-image-gen?
Run "/install qwen-image-skill" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is qwenz-image-gen free?
Yes, qwenz-image-gen is completely free (open-source). You can download, install and use it at no cost.
Which platforms does qwenz-image-gen support?
qwenz-image-gen is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created qwenz-image-gen?
It is built and maintained by irron (@irron); the current version is v1.0.0.
More Skills