返回技能目录
静态审阅软件开发

技能研究档案

detecting-ntlm-relay-with-event-correlation

来源声明

Detect NTLM relay attacks (T1557.001) by correlating Windows Event 4624 LogonType 3 for IP-to-hostname mismatches, identifying Responder/LLMNR poisoning artifacts, auditing SMB/LDAP signing, and flagging NTLMv2-to-NTLMv1 downgrades. Use for hunting credential relay in NTLM-enabled AD, investigating auth-source anomalies, building SIEM correlation rules, or responding to PetitPotam/DFSCoerce/PrinterBug alerts.

保留作者的英文原始描述。它表达设计意图,不代表已验证的实际效果。

Detect NTLM relay attacks (T1557.001) by correlating Windows Event 4624 LogonType 3 for IP-to-hostname mismatches, identifying Responder/LLMNR poisoning artifacts, auditing SMB/LDAP signing, and flagging NTLMv2-to-NTLMv1 downgrades. Use for hunting credential relay in NTLM-enabled AD, investigating auth-source anomalies, building SIEM correlation rules, or responding to PetitPotam/DFSCoerce/PrinterBug alerts.

developmentreferencesscripts

研究摘要

机器生成分析,等待人工复核

以下观察来自确定性源码解析,不代表人工安全审查或受控任务评估。

来源描述:Detect NTLM relay attacks (T1557.001) by correlating Windows Event 4624 LogonType 3 for IP-to-hostname mismatches, identifying Responder/LLMNR poisoning artifacts, auditing SMB/LDAP signing, and flagging NTLMv2-to-NTLMv1 downgrades. Use for hunting credential relay in NTLM-enabled AD, investigating auth-source anomalies, building SIEM correlation rules, or responding to PetitPotam/DFSCoerce/PrinterBug alerts. 此双语摘要由所收集的源文本和文件结构确定性地生成,等待人工编辑复核。

审阅范围

对所收集的 SKILL.md 及辅助文本/代码进行确定性静态分析。机器辅助候选和发现等待人工编辑复核;不评估动态行为或效果。

可观察能力

  • Workflow[来源陈述;暂定,未测试。证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L79]
  • Step 1: Understand NTLM Relay Attack Flow[来源陈述;暂定,未测试。证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L81]
  • Step 2: Event 4624 LogonType 3 Analysis for Relay Detection[来源陈述;暂定,未测试。证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L102]
  • Step 3: Responder Detection via Network and Event Analysis[来源陈述;暂定,未测试。证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L196]
  • Step 4: SMB Signing Enforcement Audit[来源陈述;暂定,未测试。证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L286]
  • Step 5: NTLM Downgrade Detection[来源陈述;暂定,未测试。证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L404]
  • Step 6: NTLM Audit and Restriction Policy Configuration[来源陈述;暂定,未测试。证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L495]
  • Step 7: PetitPotam and Coercion Attack Detection[来源陈述;暂定,未测试。证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L548]
  • Step 8: Build Comprehensive Correlation Dashboard[来源陈述;暂定,未测试。证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L628]

限制与范围

  • 静态分析无法评估运行时行为、命令是否成功、网络可达性或模型输出质量。
  • 未执行任何命令、脚本、网络请求、模型调用或受控技能测试。
  • 类别、标签、能力、工作流程和双语编辑摘要均为待人工复核的确定性候选,并非已验证结果。
  • 在后续管线实际执行并记录官方模型专用计数之前,无法提供精确 token 数。

适用读者

  • 评估是否检查并安装该来源技能的实践者;此受众由产品语境推断,并非测量得到的使用情况。[证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L1-L48]

技能包含哪些内容

文件清单来自记录提交中的采集源码。路径与大小属于结构观察,不代表质量评价。

个文件
6
Markdown 文件
2
个脚本
3
个引用
1
大小
103 KB
结构复杂度
中

文件清单

路径类型字节
skills/detecting-ntlm-relay-with-event-correlation/LICENSE其他11 KB
skills/detecting-ntlm-relay-with-event-correlation/references/api-reference.md引用7.2 KB
skills/detecting-ntlm-relay-with-event-correlation/scripts/agent.py脚本14.4 KB
skills/detecting-ntlm-relay-with-event-correlation/scripts/audit_smb_signing.ps1脚本13.6 KB
skills/detecting-ntlm-relay-with-event-correlation/scripts/detect_ntlm_relay.py脚本23 KB
skills/detecting-ntlm-relay-with-event-correlation/SKILL.md指令33.9 KB

SKILL.md 前置元数据

{
  "name": "detecting-ntlm-relay-with-event-correlation",
  "description": "Detect NTLM relay attacks (T1557.001) by correlating Windows Event 4624 LogonType 3 for IP-to-hostname mismatches, identifying Responder/LLMNR poisoning artifacts, auditing SMB/LDAP signing, and flagging NTLMv2-to-NTLMv1 downgrades. Use for hunting credential relay in NTLM-enabled AD, investigating auth-source anomalies, building SIEM correlation rules, or responding to PetitPotam/DFSCoerce/PrinterBug alerts.",
  "domain": "cybersecurity",
  "subdomain": "threat-hunting",
  "tags": [
    "threat-hunting",
    "NTLM-relay",
    "event-correlation",
    "T1557.001",
    "Event-4624",
    "Responder",
    "SMB-signing",
    "LDAP-signing",
    "NTLM-downgrade",
    "PetitPotam",
    "Active-Directory"
  ],
  "version": "1.0",
  "author": "mukul975",
  "license": "Apache-2.0",
  "atlas_techniques": [
    "AML.T0051",
    "AML.T0054",
    "AML.T0056",
    "AML.T0020"
  ],
  "d3fend_techniques": [
    "Application Protocol Command Analysis",
    "Network Isolation",
    "Network Traffic Analysis",
    "Client-server Payload Profiling",
    "Network Traffic Community Deviation"
  ],
  "nist_ai_rmf": [
    "MEASURE-2.7",
    "MEASURE-2.5",
    "GOVERN-6.1",
    "MAP-5.1"
  ],
  "nist_csf": [
    "DE.CM-01",
    "DE.AE-02",
    "DE.AE-07",
    "ID.RA-05"
  ],
  "mitre_attack": [
    "T1046",
    "T1057",
    "T1082",
    "T1083",
    "T1069"
  ]
}

对运行顺序的一种解读

此工作流根据源码材料重建表面顺序。除非明确记录受控评估,否则它属于暂定解读,而非真实运行轨迹。

  1. 01
    操作

    Overview

    来源描述的步骤候选,位于“Overview”;未经执行验证。

  2. 02
    判断

    When to Use

    来源描述的决策候选,位于“When to Use”;未经执行验证。

  3. 03
    操作

    Workflow

    来源描述的步骤候选,位于“Workflow”;未经执行验证。

  4. 04
    操作

    Step 1: Understand NTLM Relay Attack Flow

    来源描述的步骤候选,位于“Step 1: Understand NTLM Relay Attack Flow”;未经执行验证。

  5. 05
    操作

    Step 2: Event 4624 LogonType 3 Analysis for Relay Detection

    来源描述的步骤候选,位于“Step 2: Event 4624 LogonType 3 Analysis for Relay Detection”;未经执行验证。

  6. 06
    操作

    Step 3: Responder Detection via Network and Event Analysis

    来源描述的步骤候选,位于“Step 3: Responder Detection via Network and Event Analysis”;未经执行验证。

查看全部工作流候选节点 (29 个候选步骤)
  1. 07
    操作

    Step 4: SMB Signing Enforcement Audit

    来源描述的步骤候选,位于“Step 4: SMB Signing Enforcement Audit”;未经执行验证。

  2. 08
    操作

    Step 5: NTLM Downgrade Detection

    来源描述的步骤候选,位于“Step 5: NTLM Downgrade Detection”;未经执行验证。

  3. 09
    操作

    Step 6: NTLM Audit and Restriction Policy Configuration

    来源描述的步骤候选,位于“Step 6: NTLM Audit and Restriction Policy Configuration”;未经执行验证。

  4. 10
    操作

    Step 7: PetitPotam and Coercion Attack Detection

    来源描述的步骤候选,位于“Step 7: PetitPotam and Coercion Attack Detection”;未经执行验证。

  5. 11
    操作

    Step 8: Build Comprehensive Correlation Dashboard

    来源描述的步骤候选,位于“Step 8: Build Comprehensive Correlation Dashboard”;未经执行验证。

  6. 12
    操作

    Key Concepts

    来源描述的步骤候选,位于“Key Concepts”;未经执行验证。

  7. 13
    操作

    Tools & Systems

    来源描述的步骤候选,位于“Tools & Systems”;未经执行验证。

  8. 14
    操作

    Common Scenarios

    来源描述的步骤候选,位于“Common Scenarios”;未经执行验证。

  9. 15
    操作

    Query Event 4624 on 10.10.20.100 filtered for IpAddress=10.10.5.50: find 15 successful NTLM logons in 30 minutes from 8 different user accounts

    来源描述的步骤候选,位于“Scenario 1: Responder Poisoning with NTLM Relay to File Server”;未经执行验证。

  10. 16
    操作

    Cross-reference 10.10.5.50 with DHCP logs and DNS: host is not a registered domain member, MAC address shows a Linux-based NIC

    来源描述的步骤候选,位于“Scenario 1: Responder Poisoning with NTLM Relay to File Server”;未经执行验证。

  11. 17
    操作

    Query Zeek network logs for 10.10.5.50: identify LLMNR responses (UDP 5355) to multiple workstations and SMB connections to 10.10.20.100

    来源描述的步骤候选,位于“Scenario 1: Responder Poisoning with NTLM Relay to File Server”;未经执行验证。

  12. 18
    操作

    Confirm IP-hostname mismatch: WorkstationName values (WS-FINANCE01, WS-HR03, etc.) all resolve to different IPs in DNS, not 10.10.5.50

    来源描述的步骤候选,位于“Scenario 1: Responder Poisoning with NTLM Relay to File Server”;未经执行验证。

  13. 19
    操作

    Check SMB signing on 10.10.20.100: RequireSecuritySignature is False, enabling the relay attack

    来源描述的步骤候选,位于“Scenario 1: Responder Poisoning with NTLM Relay to File Server”;未经执行验证。

  14. 20
    操作

    Contain: block 10.10.5.50 at the switch, force password reset for all 8 affected accounts, enable SMB signing on the file server

    来源描述的步骤候选,位于“Scenario 1: Responder Poisoning with NTLM Relay to File Server”;未经执行验证。

  15. 21
    操作

    Remediate: disable LLMNR and NBT-NS via GPO, enforce SMB signing domain-wide

    来源描述的步骤候选,位于“Scenario 1: Responder Poisoning with NTLM Relay to File Server”;未经执行验证。

  16. 22
    操作

    Confirm the anomaly: DC01$ should only authenticate from 10.10.1.10, but Event 4624 shows authentication from 10.10.5.50 via NTLM (not Kerberos)

    来源描述的步骤候选,位于“Scenario 2: PetitPotam Relay to AD Certificate Services”;未经执行验证。

  17. 23
    操作

    Check for certificate enrollment: query AD CS logs for certificate requests from DC01$ around the same timestamp -- find a certificate issued for DC01$

    来源描述的步骤候选,位于“Scenario 2: PetitPotam Relay to AD Certificate Services”;未经执行验证。

  18. 24
    操作

    Identify the attack: PetitPotam coerced DC01 to authenticate to 10.10.5.50, which relayed the authentication to ADCS01 to request a certificate for DC01$

    来源描述的步骤候选,位于“Scenario 2: PetitPotam Relay to AD Certificate Services”;未经执行验证。

  19. 25
    操作

    Assess impact: with a DC certificate, the attacker can authenticate as DC01$ and perform DCSync to extract all domain credentials

    来源描述的步骤候选,位于“Scenario 2: PetitPotam Relay to AD Certificate Services”;未经执行验证。

  20. 26
    操作

    Revoke the fraudulently issued certificate immediately

    来源描述的步骤候选,位于“Scenario 2: PetitPotam Relay to AD Certificate Services”;未经执行验证。

  21. 27
    操作

    Check for DCSync activity: query Event 4662 for directory replication from non-DC sources

    来源描述的步骤候选,位于“Scenario 2: PetitPotam Relay to AD Certificate Services”;未经执行验证。

  22. 28
    操作

    Contain: isolate 10.10.5.50, revoke certificate, patch EFS (MS-EFSRPC), enforce EPA on AD CS, require LDAP signing on all DCs

    来源描述的步骤候选,位于“Scenario 2: PetitPotam Relay to AD Certificate Services”;未经执行验证。

  23. 29
    输出

    Output Format

    来源描述的步骤候选,位于“Output Format”;未经执行验证。

潜在风险面

静态发现用于提示安装前需要检查的位置,不能证明一项技能安全、恶意或有效。

建议重点关注

收集的文本中有 58 项确定性词法发现需要复核。发现可能仅位于文档中,不证明已执行或可被利用;仅含二进制的文件未扫描。

网络访问已观察到
文件写入已观察到
凭据使用未观察到
中

文本提及网络访问

17 次命中

此源文件行包含 URL 或网络客户端。应检查目标地址、发送的数据和信任边界;静态文本无法证明实际发生了网络活动。 该规则在收集文本中命中 17 处;此处列出 3 个代表位置。命中文本可能只是文档,分析器没有执行它。

提示

文本描述或实现了文件修改

39 次命中

此行包含可能的文件写入或重定向操作。应检查目标路径和覆盖行为;静态审查没有写入任何文件。 该规则在收集文本中命中 39 处;此处列出 3 个代表位置。命中文本可能只是文档,分析器没有执行它。

上下文可能消耗在哪里

发现、激活与资源层具有不同含义。只有使用指定官方方法并记录模型的运行时,才显示精确数值。

精确 token 计数不可用

在管线使用明确模型执行官方精确计数之前,包括文件清单在内的所有 token 值均为 null。

未使用要求的官方方法记录 token 测量。本档案不会用字符数、文件大小或估计值替代。

受控评估

没有受控评估

该技能仅具备以上标明的证据。不声称任何任务成功率、token 效率、延迟或质量优势。

尚未测试

在源码可见的情况下使用

安装命令
Inspect the pinned source at https://github.com/mukul975/Anthropic-Cybersecurity-Skills/blob/54a798831d2266a3ca61ce68a7acb80b81160d57/skills/detecting-ntlm-relay-with-event-correlation/SKILL.md and follow the source repository's installation instructions. No installation was performed or verified by this static analyzer. Source license: Apache-2.0.

前置条件

  • 静态分析无法评估运行时行为、命令是否成功、网络可达性或模型输出质量。
  • 未执行任何命令、脚本、网络请求、模型调用或受控技能测试。
  • 类别、标签、能力、工作流程和双语编辑摘要均为待人工复核的确定性候选,并非已验证结果。
  • 在后续管线实际执行并记录官方模型专用计数之前,无法提供精确 token 数。

示例提示词

未提供

附上限制条件的判断

优势

  • 快照包含 6 个收集文件、2 个 Markdown 文件和 3 个脚本文件,可进行源代码级检查。[证据:skills/detecting-ntlm-relay-with-event-correlation/SKILL.md:L1]

注意事项

  • 尚未确立功能、安全或模型质量结果。安装或发布前,人工复核必须检查源代码上下文和所有引用发现。
研究结论

仅为暂定的确定性静态清单;等待人工编辑复核以及独立的受控评估。不构成安装建议、安全保证或有效性声明。

从结构上保持可追溯

档案是对公开工件的限时阅读,而非背书。请通过来源链接检查确切的上游材料。

mukul975Anthropic-Cybersecurity-Skillsskills/detecting-ntlm-relay-with-event-correlation
来源类型
社区
提交
54a798831d2266a3ca61ce68a7acb80b81160d57
采集时间
2026年11月20日
最近核查
2026年11月20日
上游许可证
Apache-2.0
仓库星标
36,446
内容依据

仅对提供的文件进行确定性词法解析、YAML frontmatter 校验、文件清单、来源描述的工作流程提取和启发式暂定分类。由机器生成;无 LLM、执行、动态测试或人工编辑批准。