The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/imiva_mcp.py utilizes subprocess.Popen to execute shell commands, specifically to run npx for package interaction.
[EXTERNAL_DOWNLOADS]: The skill is designed to download the @infimind/ecom-content-cli package from the NPM registry at runtime as specified in SKILL.md and references/mcp-config.example.json.
[REMOTE_CODE_EXECUTION]: The use of npx -y in the execution flow in scripts/imiva_mcp.py results in the execution of code from a remote third-party package on the host system.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection through its data ingestion points.
Ingestion points: The images and sellingPoints arguments in the create_koc_collage_task tool (documented in SKILL.md) allow for the inclusion of external URLs and user-provided text.
Boundary markers: There are no explicit delimiters or instructions to the model to ignore potential injection patterns within the processed content.
Capability inventory: The skill has the capability to execute shell commands via subprocess.Popen in scripts/imiva_mcp.py.
Sanitization: The wrapper script does not implement sanitization or validation logic for arguments passed to the external CLI tool.

ai-ecommerce-expert-mother-baby-ecommerce-content