AI, LLM, MCP and Agent Security

Use these techniques only on systems you own or are explicitly authorized to test. Run high-impact checks in an isolated environment with synthetic data, test credentials, controlled callback services, and rollback procedures.

Start here

• Begin with AI Security Testing Methodology for the complete workflow.
• Use the LLM guides for prompt injection, retrieval, leakage, tenant isolation, and output handling.
• Use the agent guides for tool authorization, memory, multi-agent trust, autonomy, and runtime controls.
• Use the MCP guides for server inventory, tool poisoning, authorization, transport, isolation, and supply-chain security.

Useful tools

• garak – automated LLM vulnerability scanner with probes for leakage, jailbreaks, encoding, and unsafe behavior.
• PyRIT – Microsoft framework for orchestrating repeatable AI red-team attacks and scoring responses.
• promptfoo – declarative adversarial evaluations and CI regression testing for prompts, models, and RAG systems.
• MCP Inspector – enumerate resources, prompts, tools, schemas, and server responses during authorized testing.
• mcp-scan – inspect MCP configurations and servers for tool poisoning and related risks.

How to fix and retest

1. Test with explicit authorization and synthetic data.
2. Prioritize deterministic controls around the model.
3. Keep findings and regression tests versioned as the system changes.

Retest checklist

• Repeat the original proof with the same identity, state, model version, and input source.
• Try equivalent variants through every other input channel and fallback model.
• Confirm the control is enforced by the backend or runtime, not only by prompt wording.
• Check logs and alerts, then add the case to a versioned regression suite.

Conclusion

AI security issues are easiest to manage when every piece of content is treated according to its real trust level and every consequential action is controlled outside the model. Keep the test evidence reproducible, limit blast radius, and rerun the suite whenever the model, prompt, retrieval pipeline, tool, server, or policy changes.