skills.rest
blamejs
blamejsOFFICIAL
Explore all skills by this author →
WebsiteGitHub

rag-pipeline-security

Assess RAG pipeline security risks and map them to MITRE ATLAS TTPs.

Updated May 11, 2026
Search Tags:#AI security#prompt injection#rag security#vector store poisoning#compliance gaps#MITRE ATLAS#retrieval filter bypass
One-click install
npx skills add https://github.com/blamejs/exceptd-skills --skill rag-pipeline-security
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: rag-pipeline-security
Source: https://github.com/blamejs/exceptd-skills/tree/main/skills/rag-pipeline-security
Command: npx skills add https://github.com/blamejs/exceptd-skills --skill rag-pipeline-security

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Current security and compliance frameworks (including NIST 800-53, ISO 27001, and SOC 2) have no controls for RAG-specific attack surfaces such as embedding manipulation, vector store poisoning, retrieval filter bypass, and indirect prompt injection via retrieved documents. These are demonstrated, real-world threats in mid-2026 that leave RAG pipelines completely unaddressed by standard security audits and compliance requirements.

Core Features & Use Cases

  • Attack Class Mapping: Maps 5 core RAG attack classes to MITRE ATLAS v2026.06 and MITRE ATT&CK v19.1 TTPs for consistent threat tracking.
  • Compliance Gap Analysis: Explicitly flags which controls in major global frameworks (ISO, NIST, EU AI Act, UK CAF, AU Essential Eight) fail to cover RAG-specific risks, with clear explanations of why existing controls are insufficient.
  • Posture Scoring: Provides a structured 0-80 point scoring methodology to quantify RAG pipeline security maturity across 8 key control areas.
  • Defensive Mapping: Links identified RAG risks to D3FEND v1.3.0+ defensive countermeasures for actionable remediation.
  • Use Case: A security team assessing a customer support RAG system can use this skill to identify vector store poisoning risks, validate that retrieval filters are applied pre-similarity to prevent bypass, and generate auditor-ready documentation of compliance gaps for RAG-specific threats.

Quick Start

Use the rag-pipeline-security skill to assess the security posture of the internal employee knowledge base RAG system, including mapping identified attack classes to relevant MITRE ATLAS TTPs and documenting compliance framework gaps for the audit team.

Frequently Asked Questions about rag-pipeline-security

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What RAG security threats are not covered by standard compliance frameworks?▼

Standard frameworks like ISO 27001 and NIST 800-53 lack controls for RAG threats such as vector store poisoning, embedding manipulation, and retrieval filter bypass. This skill identifies these exact compliance gaps.

How do I assess RAG pipeline vulnerabilities like indirect prompt injection?▼

Assess RAG pipeline vulnerabilities by mapping attack classes like indirect prompt injection to MITRE ATLAS TTPs. This identifies specific threats and correlates them with D3FEND defensive countermeasures for remediation.

Can I use this to map RAG attacks to MITRE ATLAS and ATT&CK frameworks?▼

Yes, you can map five core RAG attack classes, including chunking exploitation, to MITRE ATLAS v2026.06 and ATT&CK v19.1 TTPs. This ensures consistent threat tracking and structured posture scoring.

How do I score the security posture of a Retrieval-Augmented Generation system?▼

Score the security posture of a Retrieval-Augmented Generation system using a structured 0-80 point methodology. This quantifies RAG pipeline security maturity across 8 key control areas for auditor-ready reporting.

Does this skill address EU AI Act compliance gaps for vector store poisoning?▼

Yes, this skill explicitly flags which controls in frameworks like the EU AI Act fail to cover vector store poisoning. It explains why existing controls are insufficient for RAG-specific risks.

Related Skills

Explore More Legal & Compliance Skills→
devCharuzudevCharuzu

rag-architect

Design and deploy production-grade RAG pipelines with chunking, embeddings, vector stores, and evaluation.

Community
Advanced
blamejsblamejs

ai-attack-surface

Assess AI/ML attack surfaces mapped to MITRE ATLAS v2026.06.

Official
Advanced
uphiagouphiago
1.1k

hunt-llm-ai

Identify exploitable LLM/AI vulnerabilities crossing trust boundaries with validation gates.

Community
Advanced
open-agreementsopen-agreements
46

iso-27001-internal-audit

Facilitates ISO 27001:2022 internal audits with a structured workflow for control assessment, evidence collection, and findings generation.

Official
Advanced