The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, creating a potential surface for indirect prompt injection attacks.
Ingestion points: The document_review task and verify_citations mode in SKILL.md explicitly ingest external text, such as contract contents or existing AI-generated responses, for analysis.
Boundary markers: The instructions do not define specific delimiters or instructions to the agent to ignore potentially malicious commands embedded within the analyzed legal text.
Capability inventory: The skill environment possesses file-writing capabilities (documented for credential storage in moai-credentials.env) and network access through the referenced MCP tools to query various government legal databases.
Sanitization: There is no evidence of input validation, filtering, or escaping of the external text before it is processed by the agent's analytical tools.
[EXTERNAL_DOWNLOADS]: The skill documentation references and relies on the korean-law-mcp (v4.4) hosted on a third-party GitHub repository (chrisryugj/korean-law-mcp). This repository provides the core logic and interface for the 42 government APIs used by the skill's tools.

legal-law-research