ref.anthropic.building-effective-agents · design-engineering
Building Effective Agents — Anthropic Engineering
Anthropic engineering article distinguishing predefined workflows from autonomous agents, recommending simple composable patterns before complex frameworks, and describing when task complexity and tool use justify greater autonomy. It is practitioner guidance for orchestration decisions, not a visual-design standard.
https://www.anthropic.com/engineering/building-effective-agents
この標準文書は未翻訳です。以下に英語の原文を表示します。
この情報源から得られる知見
- Use only to determine appropriate autonomy and orchestration levels: simple composable workflow patterns before complex agent frameworks; increase autonomy only when task complexity and tool diversity genuinely require it.
- Map workflow/agent architecture choices to visible user controls: approval boundaries, trace visibility, stop, retry, and escalation paths must be proportionate to autonomy level.
- Distinguish predefined workflow steps from agent decisions; users must be able to understand, interrupt, and recover from both.
AwesomeDSへの反映先
rule.ai.ux-agency-contract— Preserve proportionate user agency over AI actionrule.ai.ux-uncertainty-contract— Communicate evidence and uncertainty at the decision point
実装・成果物
直接リンクされた成果物はありません。
注意事項と模倣防止
Transfer orchestration level selection and visible-control proportionality principles; this source does not govern visual design, token use, or component behaviour.