faiss
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_EXECUTION]: The LangChain integration snippet in SKILL.md uses the allow_dangerous_deserialization=True flag with FAISS.load_local. This allows the use of the pickle module for deserialization, which is a known vector for arbitrary code execution if the data source is not trusted.\n- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the creation and search of vector stores for RAG workflows. It lacks instructions or examples for implementing boundary markers or sanitization of retrieved data, which could allow malicious content within the vector store to influence agent behavior.\n- [COMMAND_EXECUTION]: The documentation includes standard pip install commands, which execute shell commands to install the required Python packages on the host system.
Audit Metadata