Unverified90% confidenceFactTime unknown
System Prompt与用户输入本质上都是模型的输入token，模型并不存在独立的规则执行引擎来强制保证指令被遵守
1
Sources
90%
Confidence
Long-term
Relevance
5/31/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn Claude Code's architecture, the model cannot directly execute tool operations; it can only return instructions to the harness for processing.63% similarUnverifiedWhen using Claude Code, users do not need to pre-format their requirements or write any prompt templates; they simply input a raw problem description to the AI.62% similarUnverifiedIn Auto Mode, tool execution results are not sent to the classifier as an information isolation strategy to guard against Prompt Injection attacks.62% similarUnverified全权限模式不能通过快捷键切换，需要通过特定的命令行参数来启动61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/12956API
curl https://kongchang.com/api/v1/knowledge/claims/12956MCP
get_claim(id=12956)