The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill documents RAG implementation patterns which involve processing external data. This creates a surface for indirect prompt injection where instructions hidden in ingested documents could influence the agent's behavior.
Ingestion points: The docs parameter in the build_rag_pipeline function in SKILL.md.
Boundary markers: None present in the provided code snippet.
Capability inventory: Retrieval-based question answering using the RetrievalQA chain.
Sanitization: None present in the provided code snippet.
[SAFE]: All components and code examples utilize well-known, industry-standard libraries such as LangChain and Chroma. No patterns of data exfiltration, command execution, or obfuscation were found.

agentic-workflows