The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the shell to perform domain checks and environment verification.
Uses bash loops to iterate through candidate domains and execute vercel, whois, and curl commands.
Employs complex command pipelines involving grep, sed, awk, jq, and xmllint to parse tool outputs.
Executes vercel domains check, vercel domains price, and vercel domains inspect for registrar operations.
[EXTERNAL_DOWNLOADS]: The skill recommends installing external CLI tools to provide its functionality.
Suggests global installation of the Vercel CLI via npm i -g vercel.
Recommends installing the whois utility via brew install whois on macOS.
[REMOTE_CODE_EXECUTION]: The skill includes instructions to execute code from an untrusted third-party source.
Recommends using npx unclaimed for bulk multi-TLD sweeps. npx downloads and executes the unclaimed package from the NPM registry at runtime. The package is maintained by an individual ('iannuttall') who is not on the trusted vendors list.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, creating an attack surface for indirect prompt injection.
Ingestion points: Fetches content from arbitrary domains during liveness probes using curl (Step 7a). It also ingests search results from the USPTO website via agent-browser (Step 10a).
Boundary markers: The skill does not employ any specific delimiters or warnings to the agent to ignore instructions that might be embedded in the fetched web content.
Capability inventory: The skill has significant capabilities, including arbitrary shell command execution, network requests via curl, and browser automation via agent-browser.
Sanitization: External content is processed using basic text-processing tools like grep and head, which do not sanitize or escape potentially malicious natural language instructions before they are returned to the agent's context.

domain