The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the system whois utility through the scripts/check-domains.mjs script to verify domain registration status. The execution uses node:child_process.execFile with an input sanitization step that employs a whitelist regex /[^a-z0-9.-]/g, effectively neutralizing shell command injection attempts.
[EXTERNAL_DOWNLOADS]: The scripts/check-domains.mjs script performs network requests to rdap.org. This is a well-known and legitimate public service for accessing domain registration data, which aligns with the skill's primary function of brand and domain research.
[PROMPT_INJECTION]: The skill features an indirect prompt injection surface as it is designed to ingest and analyze external market data, competitor names, and taglines from the web.
Ingestion points: Market scan phase in SKILL.md where external competitor data is retrieved via browsing.
Boundary markers: Data is structured into markdown tables as defined in references/market-scan.md.
Capability inventory: The skill possesses the ability to execute subprocesses (whois) and perform network requests (rdap.org).
Sanitization: Candidate names derived from the workflow are sanitized via regex in the domain-checking script before being processed as command arguments.

product-brand-naming