RAG / Knowledge System Security Packet

Scope Brief

• Organization: To be specified during scoping
• Decision: A tenant-safe retrieval model with a leakage findings list and remediation backlog.
• Driver: RAG retrieval, embeddings, ingestion, and tenant boundaries may leak customer or cross-tenant data.

Authorization Statement

• Not yet confirmed — required before any access or testing.

Evidence Handling Plan

• Evidence stored in an access-controlled encrypted store; redaction; agreed retention + deletion.

Contract Requirements

• Mutual NDA — Required for this engagement.
• Evidence Handling Policy — Required for this engagement.
• Statement of Work Template — Required for this engagement.
• Assessment Terms Addendum — Required for this engagement.
• No-Cost Scoping Retainer (conditional) — Scope before any paid work or active testing.

Draft SOW Inputs

• Engagement: RAG / Knowledge System Security Packet
• Deliverables: 8 selected
• Budget category: Product security / data security / AI program

Open Questions

• Authorization (own / control / explicitly authorized to assess the targets)
• Target inventory (named systems/endpoints in scope)
• RAG boundary (sources, tenant boundaries, ingestion)
• Access plan (how access and test accounts are provisioned securely)

Follow-On Recommendations

• ai red team adversarial testing
• ai product security assessment
• claim readiness evidence

Target Inventory

• No named targets yet — required before testing.

RAG Boundary

• RAG boundary needed.

Access Plan

• Access plan needed — secure-channel provisioning, no public credentials.

Data Handling Plan

• Sensitive data in scope: No
• Standard handling; secrets masked; minimum-necessary access.

Deliverables Plan

• RAG Architecture Intake
• Data Source Inventory
• Tenant Boundary Map
• Retrieval Permission Model
• Document Ingestion Review
• Prompt Injection Test Plan
• RAG Leakage Findings
• Remediation Backlog