RAG Contamination Threat Model.

Required input fields.

• Source trust (required): Curated internal corpus, Mixed internal and external, Open web or user-submitted
• Ingestion governance (required): Reviewed before indexing, Scheduled automated ingestion, Continuous or user-driven
• Retrieval access control (required): Strict per-user access, Partial filters, Weak or unclear
• Sensitive content (required): None expected, Some controlled content, Material sensitive content
• Citation validation (required): Strict validation, Partial, None
• Downstream action (required): No downstream action, Draft-only, Workflow update, External action

Data handling: this interface uses the L2ET protected same-origin instrument engine. Do not enter confidential, regulated, privileged, incident, medical or sensitive operational data.

Pipeline logic.

Scores each RAG stage and produces a path diagram with controls for poisoning, prompt injection, stale sources, citation laundering and permission leakage.

Source families

Assumptions

• The model is architectural and depends on user-described controls.
• RAG risks vary by corpus, retrieval policy and action design.
• Testing is still required.

RAG contamination path report.

The generated artifact includes findings, assumptions, limitations, recommended next actions and exportable structured output.

Export options

Example input and output.

Example input

Mixed-source RAG with scheduled ingestion, partial access filtering, some sensitive content and draft-only output.

Example output

Produces contamination scenarios, controls, monitoring requirements, release blockers and residual-risk notes.