返回技能目录
静态审阅软件开发

技能研究档案

implementing-data-loss-prevention-with-microsoft-purview

来源声明

Implements DLP policies using Microsoft Purview PowerShell cmdlets and the Graph API to protect data across Exchange Online, SharePoint, OneDrive, Teams, endpoints, and Power BI, including sensitivity labels, custom sensitive information types with regex, endpoint DLP rules, and Activity Explorer monitoring. Use when stopping PII/PHI/PCI exfiltration, configuring sensitivity labels, or investigating DLP incidents for policy tuning.

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。

Implements DLP policies using Microsoft Purview PowerShell cmdlets and the Graph API to protect data across Exchange Online, SharePoint, OneDrive, Teams, endpoints, and Power BI, including sensitivity labels, custom sensitive information types with regex, endpoint DLP rules, and Activity Explorer monitoring. Use when stopping PII/PHI/PCI exfiltration, configuring sensitivity labels, or investigating DLP incidents for policy tuning.

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:Implements DLP policies using Microsoft Purview PowerShell cmdlets and the Graph API to protect data across Exchange Online, SharePoint, OneDrive, Teams, endpoints, and Power BI, including sensitivity labels, custom sensitive information types with regex, endpoint DLP rules, and Activity Explorer monitoring. Use when stopping PII/PHI/PCI exfiltration, configuring sensitivity labels, or investigating DLP incidents for policy tuning. 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/implementing-data-loss-prevention-with-microsoft-purview/SKILL.md:L56]
  • Step 1: Design the Sensitivity Label Taxonomy[来源陈述;暂定,未测试。证据:skills/implementing-data-loss-prevention-with-microsoft-purview/SKILL.md:L58]
  • Step 2: Create DLP Policies with Sensitive Information Types[来源陈述;暂定,未测试。证据:skills/implementing-data-loss-prevention-with-microsoft-purview/SKILL.md:L124]
  • Step 3: Deploy Endpoint DLP Rules[来源陈述;暂定,未测试。证据:skills/implementing-data-loss-prevention-with-microsoft-purview/SKILL.md:L237]
  • Step 4: Configure Auto-Labeling Policies[来源陈述;暂定,未测试。证据:skills/implementing-data-loss-prevention-with-microsoft-purview/SKILL.md:L343]
  • Step 5: Monitor with Activity Explorer and Manage DLP Alerts[来源陈述;暂定,未测试。证据:skills/implementing-data-loss-prevention-with-microsoft-purview/SKILL.md:L394]

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/implementing-data-loss-prevention-with-microsoft-purview/SKILL.md:L1-L34]

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
4
Markdown 文件
2
个脚本
1
个引用
1
大小
61.4 KB
结构复杂度
低

文件清单

路径类型字节
skills/implementing-data-loss-prevention-with-microsoft-purview/LICENSE其他11 KB
skills/implementing-data-loss-prevention-with-microsoft-purview/references/api-reference.md引用5.2 KB
skills/implementing-data-loss-prevention-with-microsoft-purview/scripts/agent.py脚本14.5 KB
skills/implementing-data-loss-prevention-with-microsoft-purview/SKILL.md指令30.7 KB

SKILL.md 前置元数据

{
  "name": "implementing-data-loss-prevention-with-microsoft-purview",
  "description": "Implements DLP policies using Microsoft Purview PowerShell cmdlets and the Graph API to protect data across Exchange Online, SharePoint, OneDrive, Teams, endpoints, and Power BI, including sensitivity labels, custom sensitive information types with regex, endpoint DLP rules, and Activity Explorer monitoring. Use when stopping PII/PHI/PCI exfiltration, configuring sensitivity labels, or investigating DLP incidents for policy tuning.\n",
  "domain": "cybersecurity",
  "subdomain": "data-protection",
  "tags": [
    "DLP",
    "Microsoft-Purview",
    "sensitivity-labels",
    "endpoint-DLP",
    "data-classification",
    "compliance"
  ],
  "version": "1.0.0",
  "author": "mukul975",
  "license": "Apache-2.0",
  "nist_csf": [
    "PR.DS-01",
    "PR.DS-02",
    "PR.DS-10",
    "GV.PO-01"
  ],
  "mitre_attack": [
    "T1486",
    "T1530",
    "T1537",
    "T1048",
    "T1573"
  ]
}

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
    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。

  2. 02
    操作

    Workflow

    来源描述的步骤候选,位于“Workflow”;未经执行验证。

  3. 03
    操作

    Step 1: Design the Sensitivity Label Taxonomy

    来源描述的步骤候选,位于“Step 1: Design the Sensitivity Label Taxonomy”;未经执行验证。

  4. 04
    操作

    Establish label tiers: Create a label hierarchy reflecting data sensitivity levels. A standard enterprise taxonomy includes:

    来源描述的步骤候选,位于“Step 1: Design the Sensitivity Label Taxonomy”;未经执行验证。

  5. 05
    操作

    Define protection settings per label: For each label, configure encryption scope (all employees, specific groups, or custom permissions), content marking (headers, footers, watermarks), and auto-labeling conditions:

    来源描述的步骤候选,位于“Step 1: Design the Sensitivity Label Taxonomy”;未经执行验证。

  6. 06
    操作

    Configure content marking: Apply visual indicators that persist with the document:

    来源描述的步骤候选,位于“Step 1: Design the Sensitivity Label Taxonomy”;未经执行验证。

查看全部工作流候选节点 (49 个候选步骤)
  1. 07
    判断

    Publish labels via label policy: Labels must be published to users through a label policy that defines which users see the labels and whether a default label or mandatory labeling is enforced:

    来源描述的决策候选,位于“Step 1: Design the Sensitivity Label Taxonomy”;未经执行验证。

  2. 08
    操作

    Step 2: Create DLP Policies with Sensitive Information Types

    来源描述的步骤候选,位于“Step 2: Create DLP Policies with Sensitive Information Types”;未经执行验证。

  3. 09
    操作

    Create a DLP policy using built-in sensitive information types: Microsoft Purview includes 300+ built-in SITs for credit card numbers, Social Security numbers, passport numbers, and health records. Create a policy targeting financial data:

    来源描述的步骤候选,位于“Step 2: Create DLP Policies with Sensitive Information Types”;未经执行验证。

  4. 10
    操作

    Create custom sensitive information types with regex: Define organization-specific patterns for data that built-in SITs do not cover:

    来源描述的步骤候选,位于“Step 2: Create DLP Policies with Sensitive Information Types”;未经执行验证。

  5. 11
    操作

    Use sensitivity labels as DLP conditions: Create policies that apply different restrictions based on the label applied to the content:

    来源描述的步骤候选,位于“Step 2: Create DLP Policies with Sensitive Information Types”;未经执行验证。

  6. 12
    操作

    Step 3: Deploy Endpoint DLP Rules

    来源描述的步骤候选,位于“Step 3: Deploy Endpoint DLP Rules”;未经执行验证。

  7. 13
    操作

    Verify device onboarding: Confirm devices are onboarded to Microsoft Purview endpoint DLP through Microsoft Intune or the local onboarding script:

    来源描述的步骤候选,位于“Step 3: Deploy Endpoint DLP Rules”;未经执行验证。

  8. 14
    操作

    Configure endpoint DLP settings: Define global settings that control which applications and file types endpoint DLP monitors:

    来源描述的步骤候选,位于“Step 3: Deploy Endpoint DLP Rules”;未经执行验证。

  9. 15
    输入

    Create endpoint-specific DLP rules: Define rules that control copy-to-USB, print, upload, and clipboard operations for sensitive content:

    来源描述的步骤候选,位于“Step 3: Deploy Endpoint DLP Rules”;未经执行验证。

  10. 16
    操作

    Configure printer groups and USB device exceptions: Allow specific printers and approved USB devices while blocking unauthorized removable media:

    来源描述的步骤候选,位于“Step 3: Deploy Endpoint DLP Rules”;未经执行验证。

  11. 17
    操作

    Step 4: Configure Auto-Labeling Policies

    来源描述的步骤候选,位于“Step 4: Configure Auto-Labeling Policies”;未经执行验证。

  12. 18
    操作

    Create auto-labeling policy for email: Automatically label inbound and outbound emails containing sensitive information:

    来源描述的步骤候选,位于“Step 4: Configure Auto-Labeling Policies”;未经执行验证。

  13. 19
    操作

    Create auto-labeling policy for SharePoint and OneDrive: Label existing files at rest that match sensitive information patterns:

    来源描述的步骤候选,位于“Step 4: Configure Auto-Labeling Policies”;未经执行验证。

  14. 20
    操作

    Simulate before enforcing: Always run auto-labeling in simulation mode first. Review the simulation results in the Microsoft Purview portal under Information Protection > Auto-labeling. The simulation shows estimated matches per location and sample content matches for validation. Only switch to enforcement mode after confirming accuracy:

    来源描述的步骤候选,位于“Step 4: Configure Auto-Labeling Policies”;未经执行验证。

  15. 21
    操作

    Step 5: Monitor with Activity Explorer and Manage DLP Alerts

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  16. 22
    操作

    Use Activity Explorer and the DLP alerts dashboard to monitor policy effectiveness and investigate incidents:

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  17. 23
    操作

    Access Activity Explorer: Navigate to Microsoft Purview portal > Data Classification > Activity Explorer. Filter by activity type "DLPRuleMatch" to see all DLP policy matches. Key columns include:

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  18. 24
    操作

    Activity timestamp and user principal name

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  19. 25
    操作

    Sensitive information type matched and confidence level

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  20. 26
    操作

    Policy and rule name that triggered

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  21. 27
    操作

    Action taken (Audit, Block, Warn with Override)

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  22. 28
    操作

    Location (Exchange, SharePoint, OneDrive, Endpoint)

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  23. 29
    操作

    File name and site URL

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  24. 30
    输出

    Analyze false positive rates: Export Activity Explorer data filtered by "Override" actions with justification text to identify rules that users frequently override. A high override rate (>20%) indicates the rule may be too aggressive or matching non-sensitive content:

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  25. 31
    输入

    Configure DLP alerts: Set up alert policies in Microsoft Purview > Data Loss Prevention > Alerts to receive notifications for high-severity matches:

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  26. 32
    操作

    Query DLP events via Microsoft Graph API: Programmatically retrieve DLP alerts and policy match details for integration with SIEM or custom dashboards:

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  27. 33
    操作

    Retrieve DLP policy match details for compliance reporting: Use the unified audit log to extract granular DLP match data including the matched content, SIT type, and confidence level:

    来源描述的步骤候选,位于“Step 5: Monitor with Activity Explorer and Manage DLP Alerts”;未经执行验证。

  28. 34
    操作

    Key Concepts

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

  29. 35
    操作

    Tools & Systems

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

  30. 36
    操作

    Common Scenarios

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

  31. 37
    操作

    Design a four-tier sensitivity label taxonomy: Public, Internal, Confidential (with sub-labels for Finance, Legal, HR), and Highly Confidential. Publish labels to all users with "Internal" as the default label and mandatory labeling enabled for email.

    来源描述的步骤候选,位于“Scenario: Protecting Financial Data Across a Multinational Organization”;未经执行验证。

  32. 38
    输出

    Create a DLP policy "PCI-DSS Financial Data Protection" scoped to all Exchange, SharePoint, OneDrive, Teams, and Endpoint locations. Configure two rules: a warning rule for 1-4 credit card numbers (notify user, allow override with justification) and a blocking rule for 5+ credit card numbers (block external sharing, generate incident report to compliance team).

    来源描述的步骤候选,位于“Scenario: Protecting Financial Data Across a Multinational Organization”;未经执行验证。

  33. 39
    输入

    Deploy endpoint DLP with rules blocking copy-to-USB and upload-to-unapproved-cloud for any file containing credit card numbers or labeled "Confidential - Finance". Allow printing with audit logging. Configure approved USB device exceptions for encrypted corporate drives by vendor/product ID.

    来源描述的步骤候选,位于“Scenario: Protecting Financial Data Across a Multinational Organization”;未经执行验证。

  34. 40
    操作

    Create auto-labeling policies that scan existing SharePoint finance sites and OneDrive locations, automatically applying "Confidential - Finance" to documents matching credit card number or bank account number SITs with confidence level 85+.

    来源描述的步骤候选,位于“Scenario: Protecting Financial Data Across a Multinational Organization”;未经执行验证。

  35. 41
    操作

    Run all policies in simulation mode for 14 days. Review Activity Explorer for false positive rates, override patterns, and unprotected sensitive content locations. Tune SIT confidence thresholds from 75 to 85 on the credit card SIT after identifying false positives from partial number sequences in meeting notes.

    来源描述的步骤候选,位于“Scenario: Protecting Financial Data Across a Multinational Organization”;未经执行验证。

  36. 42
    操作

    Switch to enforcement mode after stakeholder sign-off. Configure DLP alerts with Microsoft Sentinel integration for real-time incident correlation. Schedule monthly Activity Explorer reviews to track policy effectiveness metrics.

    来源描述的步骤候选,位于“Scenario: Protecting Financial Data Across a Multinational Organization”;未经执行验证。

  37. 43
    操作

    Create a custom sensitive information type using regex `RX-20[2-3][0-9]-[0-9]{4}` with corroborating keywords ("compound", "trial", "formulation", "assay", "efficacy") within 300-character proximity. Set primary pattern at 85% confidence and keyword-corroborated pattern at 95%.

    来源描述的步骤候选,位于“Scenario: Implementing Custom DLP for Intellectual Property Protection”;未经执行验证。

  38. 44
    操作

    Create a second custom SIT for compound identifiers using regex `CPD-[A-Z]{3}-[0-9]{5}` with keywords ("molecule", "synthesis", "pharmacokinetics") for higher confidence matching.

    来源描述的步骤候选,位于“Scenario: Implementing Custom DLP for Intellectual Property Protection”;未经执行验证。

  39. 45
    操作

    Deploy a DLP policy scoped to the Research department's Exchange distribution list, SharePoint research site collection, and research team OneDrive accounts. Block external sharing, block forwarding to non-research internal users, and generate alerts for the research compliance officer.

    来源描述的步骤候选,位于“Scenario: Implementing Custom DLP for Intellectual Property Protection”;未经执行验证。

  40. 46
    操作

    Configure endpoint DLP to prevent copy-to-USB and screen capture of documents containing the custom SITs on research department laptops. Allow printing only to approved secure printers in the research facility.

    来源描述的步骤候选,位于“Scenario: Implementing Custom DLP for Intellectual Property Protection”;未经执行验证。

  41. 47
    操作

    Create a sensitivity label "Highly Confidential - Research" with encryption restricting access to the Research security group. Configure auto-labeling to apply this label to documents matching either custom SIT.

    来源描述的步骤候选,位于“Scenario: Implementing Custom DLP for Intellectual Property Protection”;未经执行验证。

  42. 48
    操作

    Monitor Activity Explorer weekly for 30 days post-deployment. The compliance team identifies that the RX-YYYY-NNNN regex matches historical project codes in archived documents. Refine the regex to `RX-202[4-6]-[0-9]{4}` to target only active project codes and reduce false positives by 60%.

    来源描述的步骤候选,位于“Scenario: Implementing Custom DLP for Intellectual Property Protection”;未经执行验证。

  43. 49
    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 53 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
凭据使用已观察到
中

文本提及网络访问

18 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 18 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

31 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 31 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。

高

文本引用凭据或密钥材料

4 次命中

此行疑似凭据字面值,或读取敏感环境变量。请勿暴露其内容;应核实它是占位符、所需密钥,还是误提交的敏感材料。 该规则在收集文本中命中 4 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/implementing-data-loss-prevention-with-microsoft-purview/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 4 个收集文件、2 个 Markdown 文件和 1 个脚本文件,可进行源代码级检查。[证据:skills/implementing-data-loss-prevention-with-microsoft-purview/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/implementing-data-loss-prevention-with-microsoft-purview
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。