The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use shell commands, including whois, dig, and curl, to verify the status of domains and social media accounts. These commands are configured for specific availability lookups and piped to filters for data extraction.
[PROMPT_INJECTION]: The skill's workflow involves ingesting untrusted data from external sources, presenting an indirect prompt injection surface. Ingestion points: Data entering the context from registrar responses, shell command outputs, and web searches for trademark information. Boundary markers: The author has implemented a robust 'External Content Trust Boundary' in SKILL.md and references/validation.md which warns the agent to ignore any embedded directives. Capability inventory: The skill uses subprocess calls via shell (whois, curl, dig) and performs file-write operations to the .artifacts/docs/ directory. Sanitization: There are explicit instructions for the agent to extract only availability signals and discard any behavioral suggestions or directives found in HTML or search results.

product-naming