Unverified50% confidenceConceptExact time
Common variants of prompt injection include direct injection, indirect injection and jailbreak.
1
Sources
50%
Confidence
Long-term
Relevance
10/22/2026
First Seen
Sources
Can AI Agents Be Socially Engineered? Understanding Prompt Injection and Tool Permission Risks
youtubeRuleArena 攻防實驗室10/22/2026
Related Entities
Related Claims
UnverifiedCommon jailbreak techniques include role-play induction, gradual multi-turn conversation breakthroughs, encoding-level semantic obfuscation, and adversarial suffix techniques.71% similarUnverifiedPrompt injection can be divided into two types: direct injection and indirect injection, where indirect injection refers to attackers hiding malicious instructions in external content read by the agent.68% similarUnverifiedPrompt Injection refers to malicious text being embedded in content that the model will read, in an attempt to override or tamper with the original instructions.63% similarUnverified提示词注入分为直接注入和间接注入两种形式，因语言模型无法区分数据与指令，凡进入上下文窗口的文本都可能被执行61% similarUnverifiedThe essence of prompt injection is that attackers hijack the execution logic of large language models through carefully crafted text content.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1085883API
curl https://kongchang.com/api/v1/knowledge/claims/1085883MCP
get_claim(id=1085883)