The Agent Skills Directory

[COMMAND_EXECUTION]: The helper script scripts/imiva_mcp.py utilizes the subprocess.Popen function to invoke the npx package runner. This execution is used solely to bridge communication with the underlying Node.js-based MCP server via standard I/O.
[REMOTE_CODE_EXECUTION]: The skill is configured to download and execute the @infimind/ecom-content-cli package from the NPM registry using npx. This is the documented and official method for accessing the enterprise e-commerce tools provided by Beijing Infimind Technology Co., Ltd.
[CREDENTIALS_SAFE]: The skill correctly manages sensitive authentication data by instructing users to provide the MCP_TOKEN via environment variables rather than hardcoding credentials or storing them in plain text files within the skill repository.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface for untrusted data through the prompt and referenceAudios parameters in the create_video_generation_task tool. However, the risk is minimized as the skill primarily generates creative content and does not grant the agent high-privilege system access or the ability to perform sensitive file operations based on that input.

ai-ecommerce-expert-audio-reference-video