Unverified50% confidenceFactExact time
In agent scenarios, data security boundaries specifically involve the flow paths of data through prompts, context windows, external tool calls, and logs.
1
Sources
50%
Confidence
Long-term
Relevance
10/17/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAny scenario in which instructions and untrusted data share the same context (such as prompt injection, RAG, or agents) carries the risk that the model may follow information it was told to ignore.71% similarUnverifiedIn the reporting Agent, passing dataset IDs between tools instead of having the LLM recite raw records line by line saves tokens and reduces errors.70% similarUnverifiedAgents cannot naturally distinguish between data and instructions, which makes prompt injection threats particularly prominent in scenarios where agents browse the web or process user-uploaded content.70% similarUnverified工具调用引入了调用顺序、权限验证、数据访问范围等过程性安全因素，这些在最终输出中往往不可见69% similarUnverified合成数据在长尾场景补充和隐私敏感场景替代两个维度上价值最高，而非作为主体数据源67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1051427API
curl https://kongchang.com/api/v1/knowledge/claims/1051427MCP
get_claim(id=1051427)