AI Guardrails & Defense-in-Depth

Overview

The ai-guardrails skill provides a comprehensive framework for securing autonomous LLM agents. As agents interact with untrusted external data sources and perform actions on behalf of users, the attack surface expands significantly. This skill focuses on implementing multi-layered defenses to ensure agent safety, reliability, and integrity in live environments.

Threat Model

Autonomous agents face unique security challenges that require a proactive defense strategy:

1. Direct Prompt Injection (DPI): Malicious input explicitly attempting to hijack the agent's core instructions.
2. Indirect Prompt Injection (IPI): Injected instructions hidden in retrieved web pages, emails, or files that the agent processes.
3. Prompt Leaking: Attempts to extract system instructions, internal reasoning paths, or proprietary data.
4. Data Exfiltration: Unauthorized access and transfer of PII or sensitive internal data to external endpoints.
5. SSRF (Server-Side Request Forgery) via Tool Use: Leveraging agent-enabled tools to probe internal networks or access forbidden resources.
6. Agent Loop Exploitation: Crafting inputs that force the agent into infinite loops or resource-exhaustion scenarios.
7. Tool Abuse: Circumventing intended tool usage patterns to execute malicious commands.