The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The check-domains.sh script fetches the official Top-Level Domain (TLD) list from IANA (data.iana.org) to ensure registration checks are only performed on valid, delegated extensions. This is a trusted source for domain metadata.
[COMMAND_EXECUTION]: The skill utilizes standard system utilities including curl, dig, and whois to perform registration lookups. These tools are executed as subprocesses using double-quoted variables and proper job control, preventing command injection from user-supplied brand names.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted registration data and trademark search results from external sources.
Ingestion points: Registration records from whois, dig, and rdap.org are ingested and processed by the agent.
Boundary markers: The instructions guide the agent to present findings in a structured table format, which helps maintain context and reduces the risk of the agent following instructions embedded in registration records.
Capability inventory: Capabilities are limited to network-based name lookups and local DNS queries.
Sanitization: Brand names are URL-encoded in tm-links.sh using a robust urlenc function, and shell variables in check-domains.sh are properly quoted to ensure data is treated as strings rather than code.

name-availability