rag-security-architect

What problem does it solve?

This Skill helps you design and review retrieval-augmented generation systems so that users only retrieve documents they are authorized to see, with tenant boundaries and document-level access enforced before results are returned.

Core Features & Use Cases

• Retrieval-time authorization: Ensures tenant scope and ACL checks happen inside the vector search query, not after the store has already exposed candidate results.
• Vector-store tenant scoping: Reviews namespace, collection, or per-tenant index strategies and checks that tenant isolation cannot be bypassed by a client-supplied or missing filter.
• Embedding-risk analysis: Evaluates inversion, membership inference, poisoned documents, stale permissions, deletion handling, and citation safety for RAG systems.
• Use Case: A shared semantic search index for customer documents needs to prevent tenant A from surfacing tenant B's content while still supporting role-based access and secure citations.

Quick Start

Ask the AI to review your RAG pipeline for retrieval-time authorization, tenant-scoped vector search, document ACL propagation, and cross-tenant negative tests.