The Agent Skills Directory

[SAFE]: The skill serves as a technical resource for RAG implementation. It utilizes standard libraries and follows best practices for semantic search and retrieval workflows. No suspicious network activities or unauthorized data access patterns were found.
[PROMPT_INJECTION]: The architecture describes the ingestion of external documents and user queries, which establishes a surface for indirect prompt injection.
Ingestion points: External text data is processed for chunking and indexing in SKILL.md and references/chunking-strategies.md.
Boundary markers: Templates do not explicitly define delimiters or 'ignore' instructions for untrusted content.
Capability inventory: The skill uses network-based LLM and embedding services and performs operations on vector databases.
Sanitization: Sanitization of external content is not explicitly shown in the reference code; users should implement appropriate validation.

rag-architect