The Agent Skills Directory

[SAFE]: Trusted Service Integration
The skill integrates with established cloud service providers including ByteDance's Ark (Volcengine), Alibaba's Dashscope, and Google's Gemini.
All network interactions are limited to the official API endpoints of these well-known technology companies.
[SAFE]: Secure Credential Management
API credentials are managed by reading the user's local ~/.env file, which is a standard practice for development environments.
The script specifically retrieves only the necessary *_API_KEY and *_IMAGEGEN_MODEL variables and does not exfiltrate them to unauthorized third parties.
[SAFE]: Configuration Verification
The included 'probe' command allows users to verify their environment setup safely.
Probe output identifies available providers and models but does not expose the actual API key values.
[SAFE]: Scoped File Operations
Local file system operations are restricted to reading the configuration file and saving generated images to designated project directories.
The script uses the Pillow library to process and save images, providing a layer of validation for the returned data format.
[SAFE]: Environment Configuration
The skill provides instructions to source local shell profiles (e.g., .zshrc, .zprofile) to ensure environment variables are correctly loaded into the process.
This is a standard method for accessing user-defined configuration in command-line tools.
[SAFE]: Standard Dependencies
The skill utilizes widely-used and maintained Python packages such as google-genai, openai, pillow, and requests.
No suspicious patterns such as dynamic code execution or privilege escalation were identified.

threejs-image-generator