Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
HY Image Generation
v1.0.2腾讯云混元生图 3.0，文生图 / 图生图，智能生成贴合描述的图片。Tencent Cloud Hunyuan Image Generation 3.0, text-to-image / image-to-image, intelligently generates images matching the des...
⭐ 1· 341·3 current·3 all-time
by败毒@neck-cn
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
Name/description target Tencent Cloud Hunyuan image generation, and the code and SKILL.md require Tencent Cloud API credentials (TENCENTCLOUD_SECRET_ID and TENCENTCLOUD_SECRET_KEY). However the registry metadata declares no required environment variables or primary credential — an inconsistency that could mislead users about what secrets are needed.
Instruction Scope
SKILL.md instructs the agent to execute scripts automatically (zero-interaction) and not ask the user for confirmation. The scripts read TENCENTCLOUD_* environment variables and call aiart.tencentcloudapi.com to submit and poll jobs. The instructions also mandate returning result image URLs and warn not to fabricate results. The actions stay within the stated purpose (submit/poll image jobs) but the enforced 'execute without confirmation' behavior increases risk if run in an environment with sensitive credentials.
Install Mechanism
There is no declared install spec, but each script contains ensure_dependencies() that runs pip install tencentcloud-sdk-python via subprocess.check_call at runtime. This is expected to obtain the official SDK but does cause packages to be installed on first run; it's not using unusual or third-party URLs, but runtime installs can change the environment and should be noted.
Credentials
The only environment access the code performs is to TENCENTCLOUD_SECRET_ID, TENCENTCLOUD_SECRET_KEY and optional TENCENTCLOUD_TOKEN — these are exactly the credentials needed to call the Tencent Cloud API and are proportionate. The problem is the published skill metadata did not declare these required env vars, creating an information mismatch that could lead users to unknowingly provide credentials to a skill they didn't realize required them.
Persistence & Privilege
The skill is not set always:true and does not modify other skills or system configuration. It performs network calls to Tencent Cloud and installs a Python package locally at runtime, which are normal for this functionality and within expected privilege scope.
What to consider before installing
What to consider before installing:
- This skill requires your Tencent Cloud API keys (TENCENTCLOUD_SECRET_ID and TENCENTCLOUD_SECRET_KEY) even though the registry metadata omits them — verify and be comfortable providing those credentials.
- The SKILL.md enforces 'zero-interaction': the agent will run the included Python scripts automatically and will perform network calls to aiart.tencentcloudapi.com and may install the tencentcloud-sdk-python package at first run. If you run this in an environment with other sensitive credentials or on a shared machine, isolate it (e.g., in a VM or container).
- Review the included scripts (they are small and readable) to confirm they only use the stated env vars and endpoints; the code here only contacts Tencent Cloud and does not exfiltrate to unknown endpoints.
- If you want safer behavior, request or modify the skill to: (1) declare the required env vars in its metadata, (2) prompt the user for explicit confirmation before executing (disable zero-interaction), and (3) optionally avoid auto-installing packages or document that install will occur.
- Use least-privileged API keys for this purpose, monitor API usage/billing on your Tencent Cloud account, and consider rotating keys after testing.Like a lobster shell, security has layers — review code before you run it.
latestvk97847vdv474t2dw8cnnfw1t7n8357ab
License
MIT-0
Free to use, modify, and redistribute. No attribution required.