langchain-enterprise-rbac
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions involve processing user-provided queries for retrieval-augmented generation (RAG), which creates an attack surface for indirect prompt injection.
- Ingestion points: User queries are processed in the
retrievefunction in SKILL.md and references/retriever-per-request.md. - Boundary markers: The skill uses a placeholder prompt in examples, assuming the implementer will provide specific templates.
- Capability inventory: The skill is configured with
Bash(python:*)tools and executes LLM calls via LangChain runnables. - Sanitization: The skill includes specific mitigation patterns in references/role-scoped-tool-allowlist.md, such as regex-based denylists for SQL and shell commands to prevent unauthorized tool use.
- [EXTERNAL_DOWNLOADS]: The skill references standard dependencies required for its implementation patterns.
- References well-known packages including
langchain-core,langgraph,pinecone-client, andchromadbwhich are essential for the described functionality.
Audit Metadata