The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of an external NPM package (korean-law-mcp) and suggests cloning a repository from GitHub (github.com/chrisryugj/korean-law-mcp). These external sources are not part of the verified vendor's infrastructure and could contain unvetted code.
[COMMAND_EXECUTION]: The setup instructions include commands like npx, npm install, and git clone. Executing these commands results in the download and execution of scripts from external, unverified sources.
[PROMPT_INJECTION]: The tools verify_citations and action_plan ingest arbitrary text strings from the user to perform legal analysis. This represents an indirect prompt injection surface where an attacker could provide text containing instructions designed to manipulate the agent's subsequent reasoning.
Ingestion points: userText parameter in verify_citations and situation parameter in action_plan.
Boundary markers: Absent. The tools do not specify the use of delimiters or instructions to ignore embedded commands within the user-provided text.
Capability inventory: The skill performs network operations to fetch data from the Korean Ministry of Government Legislation APIs.
Sanitization: No evidence of input validation or sanitization is provided for the text fields processed by the tools.

korean-law-mcp