Unverified50% confidenceCautionExact time
External text reaching the model (emails, web pages, webhooks) may carry injected instructions targeting the model, so model output should never directly trigger deletions, sends, or payments—a whitelist or code check should be placed in between.
1
Sources
50%
Confidence
Long-term
Relevance
9/28/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe tool invocation requests output by a model are essentially still text; actual network requests, database queries, or code execution are always carried out by an external Runtime.69% similarUnverified提示词注入可分为直接注入和间接注入，间接注入通过被模型读取的外部文档、网页等传递指令66% similarUnverifiedChain-of-Thought提示只能在模型内部进行符号推演，无法获取实时信息、执行计算或操作文件系统66% similarUnverifiedExternal message ingestion supports two modes: direct mode (forwarded as-is) and AI Q&A mode (messages are first processed through a text model before being forwarded).64% similarUnverifiedIn the "LLM + tool calling" paradigm, the model does not directly manipulate files itself, but instead outputs structured instructions, with deterministic code handling the actual execution.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/955878API
curl https://kongchang.com/api/v1/knowledge/claims/955878MCP
get_claim(id=955878)