Unverified50% confidenceFactExact time
In systems with tools attached, if the model can call APIs, send messages, write to databases, or execute code, injected instructions that get executed may cause irreversible consequences.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedAfter the client receives the tool execution result, it needs to construct a message with the role set to tool, carrying the corresponding tool_call_id and execution result to send back to the model.73% similarUnverified工具调用机制中，模型不会直接执行任何代码，而是返回结构化的调用请求，由本地程序执行后再回传结果；因此安全边界掌握在开发者手里，模型只能请求，不能擅自读取文件或发送邮件。73% similarUnverified大模型API版本迭代可能伴随输出格式变化、系统提示词响应行为漂移及函数调用/结构化输出接口规范调整，即'行为漂移'现象71% similarVerifiedThe tool-calling (Function Calling) mechanism allows the model to invoke external tools during reasoning, such as file read/write operations, terminal commands, and search APIs.70% similarUnverifiedAfter receiving a request, the model may return a structured tool call rather than directly providing an answer.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1069269API
curl https://kongchang.com/api/v1/knowledge/claims/1069269MCP
get_claim(id=1069269)