diff --git a/common/websocket/static/aigdocs/docs/agent-scan_openSource.md b/common/websocket/static/aigdocs/docs/agent-scan_openSource.md
index 5b9dad42..55f4207a 100644
--- a/common/websocket/static/aigdocs/docs/agent-scan_openSource.md
+++ b/common/websocket/static/aigdocs/docs/agent-scan_openSource.md
@@ -176,8 +176,9 @@ Agent 安全扫描基于红队即服务（RTaaS）驱动的 Agent 能力，将
 
 - **分类体系**：采用 OWASP Top 10 for Agentic Applications 标准，对发现的漏洞进行统一归类与描述。后续将持续引入更多权威安全标准（如 CWE、MITRE ATT&CK 等），不断完善分类体系。
 - **检测维度**：系统内置多类检测能力，扫描时会根据目标 Agent 的能力进行智能探测，仅在存在对应能力时执行相应检测，以减少无效请求。主要检测类型包括：
-  - **数据安全类**：数据泄露检测（系统提示词暴露、凭据泄露、PII 泄露等）
-  - **注入攻击类**：间接注入检测（RAG 内容注入、文档注入、网页内容注入等）
+  - **数据安全类**：数据泄露检测（系统提示词暴露、凭据泄露、PII 泄露等）；硬编码密钥检测（代码或配置中的敏感凭证等）
+  - **注入攻击类**：间接注入检测（RAG 内容注入、文档注入、网页内容注入等）；直接注入检测（用户输入中的直接提示注入攻击等）
   - **权限控制类**：授权绕过检测（角色越权、管理功能滥用、多用户数据访问等）
-  - **工具滥用类**：工具滥用检测（SSRF、命令执行、文件操作、网络请求滥用等）
+  - **工具滥用类**：工具滥用检测（SSRF、命令执行、文件操作、网络请求滥用等）；文件路径遍历检测（通过路径操纵访问越权文件等）
+  - **记忆安全类**：记忆投毒检测（通过污染 Agent 长期记忆注入恶意指令等）
 
diff --git a/common/websocket/static/aigdocs/docs/agent-scan_openSource_en.md b/common/websocket/static/aigdocs/docs/agent-scan_openSource_en.md
index e6514945..192b3e30 100644
--- a/common/websocket/static/aigdocs/docs/agent-scan_openSource_en.md
+++ b/common/websocket/static/aigdocs/docs/agent-scan_openSource_en.md
@@ -171,7 +171,8 @@ Agent Security Scan leverages Red Teaming as a Service (RTaaS) driven Agent capa
 
 - **Classification System**: Adopts the OWASP Top 10 for Agentic Applications standard to classify and describe discovered vulnerabilities. The system will continue to incorporate more authoritative security standards (such as CWE, MITRE ATT&CK) to continuously improve the classification system.
 - **Detection Dimensions**: The system features built-in detection capabilities across multiple categories. During scanning, it intelligently probes based on the target Agent's capabilities, executing specific tests only when relevant capabilities are present to minimize ineffective requests. Major detection types include:
-  - **Data Security**: Data leakage detection (system prompt exposure, credential leakage, PII leakage, etc.).
-  - **Injection Attacks**: Indirect injection detection (RAG content injection, document injection, web content injection, etc.).
+  - **Data Security**: Data leakage detection (system prompt exposure, credential leakage, PII leakage, etc.); hardcoded secret detection (sensitive credentials embedded in code or config, etc.).
+  - **Injection Attacks**: Indirect injection detection (RAG content injection, document injection, web content injection, etc.); direct injection detection (direct prompt injection attacks via user input, etc.).
   - **Access Control**: Authorization bypass detection (privilege escalation, administrative function abuse, multi-user data access, etc.).
-  - **Tool Abuse**: Tool misuse detection (SSRF, command execution, file operations, network request abuse, etc.).
+  - **Tool Abuse**: Tool misuse detection (SSRF, command execution, file operations, network request abuse, etc.); file path traversal detection (unauthorized file access via path manipulation, etc.).
+  - **Memory Security**: Memory poisoning detection (injecting malicious instructions into the Agent's long-term memory, etc.).