Unverified50% confidenceCautionExact time
Guardrails based on keyword blocking can be evaded by users simply by rephrasing their expressions.
1
Sources
50%
Confidence
Long-term
Relevance
10/9/2026
First Seen
Sources
Related Entities
Related Claims
Unverified用户可以通过提示注入和语义改写等手段绕过内容检测，使实际拦截效果是概率性的而非绝对的79% similarUnverifiedMost guardrails are designed at the granularity of individual messages, checking whether each input contains violating keywords or directly requests restricted information.79% similarUnverifiedThe guardrail performs pattern scanning and semantic filtering on input content at the before-Model node and is the key line of defense against prompt injection attacks.77% similarUnverifiedAuto Mode provides allow, soft deny, and hard deny fields as guidance for the classifier to reference at each check, rather than as hard rules.70% similarUnverifiedMinimal mode strips away most formatting instructions and safety guardrail prompts, exposing the model more directly to user input.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/992462API
curl https://kongchang.com/api/v1/knowledge/claims/992462MCP
get_claim(id=992462)