Unverified50% confidenceFactExact time
In agent scenarios, minor wording differences in a prompt can lead to completely different behavioral chains, because agents rely on the prompt to decide which tool to call next and whether the task has been completed.
1
Sources
50%
Confidence
Long-term
Relevance
9/16/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedAn agent's behavioral chain is essentially a probabilistic reasoning process in which the same input may produce different tool-call sequences, which is what makes agent evaluation difficult.79% similarUnverified将所有外部能力都做成Tools写在Agent Prompt里会导致模型选工具时行为不稳定77% similarUnverified模型调用技能带有一段description，会进入Agent的上下文中，Agent可自行判断是否调用该技能77% similarUnverified不同Agent框架有不同的技能定义方式和接口规范，存在技能碎片化问题77% similarUnverified如果任务本质上是模式匹配或规则应用（可用if-else或规则引擎覆盖），使用Agent就是过度设计76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/926532API
curl https://kongchang.com/api/v1/knowledge/claims/926532MCP
get_claim(id=926532)