The Agent Skills Directory

[SAFE]: The skill provides educational content and code templates for building RAG systems using well-known services and libraries.
[SAFE]: No malicious patterns such as obfuscation, remote code execution, privilege escalation, or unauthorized data access were detected.
[PROMPT_INJECTION]: The skill describes patterns for handling untrusted external data (Indirect Prompt Injection surface). It includes recommended mitigations such as using boundary markers (CONTEXT/QUESTION headers) and explicit grounding instructions to the LLM to minimize risks associated with processing external content.

rag-patterns