langchain-pro
Pass
Audited by Gen Agent Trust Hub on May 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides standard, well-documented patterns for RAG pipelines and stateful workflows.
- [PROMPT_INJECTION]: The RAG example template interpolates external data into a prompt, which is an inherent characteristic of the technology.
- Ingestion points: The
{context}and{question}variables in the prompt template withinSKILL.md. - Boundary markers: No explicit delimiters are used in the basic example provided.
- Capability inventory: The provided code snippets demonstrate LLM invocation and local vector store retrieval, but do not include dangerous system-level operations.
- Sanitization: The example does not demonstrate input sanitization, which is expected for a high-level architectural template.
Audit Metadata