The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The script scripts/imiva_mcp.py executes @infimind/ecom-content-cli@latest via the npx -y command. This pattern downloads and runs code from the NPM registry at runtime. The use of the @latest tag creates a supply chain risk, as it automatically executes the most recent version of the package without pinning or integrity verification.
[COMMAND_EXECUTION]: The skill uses subprocess.Popen in scripts/imiva_mcp.py to invoke the npx binary. While the command is structured as a list to avoid shell injection, it still represents the execution of external binaries guided by skill logic.
[EXTERNAL_DOWNLOADS]: The skill relies on fetching the @infimind/ecom-content-cli package from the public NPM registry. This dependency is not from a source recognized as a trusted vendor or well-known service within the analysis framework.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its handling of untrusted data.
Ingestion points: User-supplied image paths and text prompts are ingested through the create_smart_refine_task tool parameters.
Boundary markers: The instructions do not define clear delimiters or warnings to prevent the AI from following malicious instructions embedded in input data.
Capability inventory: The skill possesses command execution capabilities through subprocess.Popen in scripts/imiva_mcp.py.
Sanitization: There is no evidence of explicit sanitization or validation of the input prompts or image metadata before they are passed to the downstream model.

ai-ecommerce-expert-qwen-image-3-pro-ecommerce-image