The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill requires the installation and execution of an external NPM package gangnamunni-clinic-search. It specifically suggests using npx, which downloads and runs code from the NPM registry. This poses a risk if the package name is hijacked, typosquatted, or if the package contains malicious logic.
[COMMAND_EXECUTION]: The instructions include a CLI workflow that involves executing shell commands (npx gangnamunni-clinic-search) to retrieve data. Using user-supplied keywords as arguments for CLI tools can lead to command injection if not properly sanitized by the underlying package.
[EXTERNAL_DOWNLOADS]: The skill performs HTTP requests to www.gangnamunni.com to fetch search results and process Next.js payloads. While the domain is a known service, fetching and parsing external HTML content introduces external data into the agent's context.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from a third-party website.
Ingestion points: External data enters the agent context via the __NEXT_DATA__ payload on the Gangnam Unni search results page, specifically the props.pageProps.hospitals array.
Boundary markers: The skill instructions do not specify any delimiters or instructions to ignore embedded commands within the fetched data.
Capability inventory: The agent has the capability to execute shell commands via npx and perform network operations.
Sanitization: There is no evidence of sanitization, filtering, or validation of the fetched web content before it is parsed and presented to the user.

gangnamunni-clinic-search