Slide 52
Slide 52 text
03 | 抽象がもう一段進む
再確認 — "Harness" とは
広義では Agent = Model + Harness — Harness = モデル以外の実行系全体
• LangChain — "A harness is every piece of code, configuration, and execution logic that isn't the model itself."
• Martin Fowler / B. Böckeler — "'harness' has emerged as a shorthand to mean everything in an AI agent except the model
itself."
• Databricks — "the software infrastructure that wraps around a LLM and enables it to act on tasks."
• Microsoft — "scaffolding that turns a language model into an agent that can actually do things."
• (狭義) Hugging Face — "the execution layer inside the agent: it calls the model, handles its tool calls, decides when to stop."
© LayerX Inc.
https://www.langchain.com/blog/the-anatomy-of-an-agent-harness / https://martinfowler.com/articles/harness-engineering.html /
https://www.databricks.com/blog/ai-harness / https://learn.microsoft.com/en-us/agent-framework/agents/harness
https://huggingface.co/blog/agent-glossary
52