The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill uses the npx command to download and execute the @infimind/ecom-content-cli package from the NPM registry at runtime. This is the official tool for the IMIVA platform and is used as intended for the skill's primary purpose.
[COMMAND_EXECUTION]: The script scripts/imiva_mcp.py uses subprocess.Popen to launch the MCP server. It follows security best practices by passing arguments as a list rather than a shell string, effectively preventing command injection attacks.
[PROMPT_INJECTION]: The skill has a standard attack surface for indirect prompt injection by processing external data to generate content.
Ingestion points: User-provided productImages and extraDescription strings defined in SKILL.md.
Boundary markers: The skill uses structural JSON-RPC message boundaries to separate data from instructions.
Capability inventory: The skill can execute the create_detail_page_task command to generate ecommerce infographics.
Sanitization: The Python script uses json.dumps to ensure user input is correctly escaped and formatted before being passed to the tool.
[CREDENTIALS_SAFE]: The skill manages authentication tokens securely. It instructs users to store their MCP_TOKEN in environment variables rather than hardcoding them, and the provided scripts read these credentials from the environment at runtime.

ai-ecommerce-expert-ecommerce-infographic-set