The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a wrapper script to execute a bundled runtime (assets/filtalgo-cli.cjs) located within the skill directory to interface with the vendor's commerce services.
[INDIRECT_PROMPT_INJECTION]: The skill ingests live product data from an external catalog; potential risks are mitigated by explicit instructions to treat all external data as untrusted and not as agent instructions.
[SAFE]: A comprehensive URL security policy in scripts/url-policy.js restricts output to HTTPS and prevents access to local or private network addresses, protecting against SSRF and internal data exfiltration.
[SAFE]: Privacy-preserving measures are used throughout, including masking order IDs and phone numbers and ensuring account tokens are kept private.

filtmall-shopping