The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads and executes the @infimind/ecom-content-cli@latest package from the npm registry using npx. This is the documented and expected method for interacting with the IMIVA platform's Model Context Protocol (MCP) server. The package is associated with the official vendor described in the skill content.
[COMMAND_EXECUTION]: The script scripts/imiva_mcp.py uses subprocess.Popen to invoke the node-based CLI tool. The command construction is static, and user-supplied arguments are passed through stdin as serialized JSON, which effectively prevents shell command injection.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data (image paths and product descriptions) that could contain malicious instructions.
Ingestion points: User-provided image URLs/paths and task descriptions passed to the create_visual_migration_task tool.
Boundary markers: None implemented; the skill assumes the external API handles content safety.
Capability inventory: Subprocess execution for CLI tools and network communication with the IMIVA API.
Sanitization: Arguments are validated as JSON objects and serialized using json.dumps before processing.
[CREDENTIALS_SAFE]: The skill requires an MCP_TOKEN for authentication, which is correctly managed via environment variables rather than being hardcoded. Instructions for token creation point to the official imiva.ecpro.com domain.

ai-ecommerce-expert-multi-pair-visual-migration