APPENDIX F: FURTHER READING AND RESOURCES

STANDARDS AND OFFICIAL DOCUMENTATION Resource: Model Context Protocol website and specification. Where to find it: modelcontextprotocol.io, with the source materials in the modelcontextprotocol organization on GitHub. Why read it: it is the official home of the open standard taught in Chapters 16, 39, and 67, and it is the place to check which specification version is current. Resource: Language Model Evaluation Harness by the EleutherAI research group. Where to find it: github.com/EleutherAI/lm-evaluation-harness Why read it: it is an open-source toolkit for testing language models on many standard benchmarks, used as an example in Chapter 34. Resource: Artificial Intelligence Risk Management Framework from the United States National Institute of Standards and Technology. Where to find it: nist.gov/itl/ai-risk-management-framework Why read it: it is voluntary guidance for managing risks of AI systems, released in January 2023, and it supports the compliance topics in Chapter 41. Resource: Generative AI Profile from the same institute, numbered NIST-AI-600-1. Where to find it: linked from the framework page above. Why read it: it was released in July 2024 and applies the framework to risks that are special to generative AI. Resource: OWASP GenAI Security Project, which hosts the Top 10 list of risks for Large Language Model applications. Where to find it: genai.owasp.org Why read it: it names the most common security risks, including prompt injection, and supports Chapters 38 and 39. Check which edition of the list is current, because more than one edition exists.