Unverified50% confidenceFactExact time
Traditional EDR security tools cannot identify prompt attack payloads embedded in Markdown files.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/12/2026
First Seen
Valid until: 12/11/2026
Sources
Related Entities
Related Claims
Unverified现有文档安全扫描工具主要检查宏代码、嵌入对象和外部链接等已知威胁向量，几乎没有工具会审查空格字符的渐变色标点参数是否携带编码数据58% similarUnverifiedAmazon Kiro suffered a prompt injection attack in which attackers used malicious content to cause the IDE plugin's LLM backend to read and exfiltrate developers' API keys.57% similarUnverifiedEDR系统持续记录终端设备行为事件并通过行为分析引擎识别异常，其核心价值在于应对无文件攻击和零日漏洞利用等传统防护无法识别的威胁57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/911469API
curl https://kongchang.com/api/v1/knowledge/claims/911469MCP
get_claim(id=911469)