Back to Details

internal-investigation

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Internal Investigation Skill

内部调查Skill

Matter context

Matter上下文

Matter context. Check 
## Matter workspaces
in the practice-level CLAUDE.md. If 
Enabled
is 
✗
(the default for in-house users), skip the rest of this paragraph — skills use practice-level context and the matter machinery is invisible. If enabled and there is no active matter, ask: "Which matter is this for? Run 
/employment-legal:matter-workspace switch <slug>
or say 
practice-level
." Load the active matter's 
matter.md
for matter-specific context and overrides. Write outputs to the matter folder at 
~/.claude/plugins/config/claude-for-legal/employment-legal/matters/<matter-slug>/
. Never read another matter's files unless 
Cross-matter context
is 
on
.
Matter上下文。请查看实践层级CLAUDE.md中的
## Matter workspaces
部分。如果
Enabled
为
✗
（内部用户默认设置），则跳过本段剩余内容——Skill将使用实践层级上下文，matter相关机制不可见。如果该功能已启用且当前无活跃matter，请询问：“这是针对哪个matter的？请执行
/employment-legal:matter-workspace switch <slug>
命令，或告知‘实践层级’。”加载活跃matter的
matter.md
文件以获取matter专属上下文和配置覆盖。将输出内容写入matter文件夹，路径为
~/.claude/plugins/config/claude-for-legal/employment-legal/matters/<matter-slug>/
。除非
Cross-matter context
处于开启状态，否则不得读取其他matter的文件。

Output header

输出页眉

Prepend the work-product header from 
~/.claude/plugins/config/claude-for-legal/employment-legal/CLAUDE.md
→ 
## Outputs
(it differs by user role — see 
## Who's using this
). Every file, log, memo, and summary produced by this skill opens with that header.
Distribution discipline. Every file this skill creates — log entries, memo drafts, audience summaries, document notes — inherits the privilege and confidentiality status of the underlying investigation. Distribution beyond the privilege circle (forwarding to non-attorneys outside the investigation team, cc'ing HR without scoping, handing to the business side) can waive privilege over the entire investigation. Store these files where privileged materials live, label per the work-product header, and make every distribution decision deliberately.
在本Skill生成的所有工作成果前添加
~/.claude/plugins/config/claude-for-legal/employment-legal/CLAUDE.md
中
## Outputs
部分指定的工作成果页眉（该页眉因用户角色而异——详见
## Who's using this
部分）。本Skill生成的每一份文件、日志、备忘录和摘要都需以此页眉开头。
分发规范。本Skill创建的所有文件——日志条目、备忘录草稿、受众摘要、文档注释——均继承底层调查的特权和保密状态。将文件分发至特权圈外（转发给调查团队外的非律师人员、未限定范围抄送HR部门、提交给业务部门）可能导致整个调查的特权被放弃。请将这些文件存储在特权材料专属位置，按照工作成果页眉标注，并审慎做出每一项分发决策。

⚠️ Privilege notice — read before proceeding

⚠️ 特权须知——开始前请阅读

Marking does not create privilege. The header above reflects the intended protection and is important to include — but it does not itself establish privilege. Whether any given output is actually privileged depends on whether the investigation is attorney-directed, the purpose for which documents are created, and how they are subsequently used or disclosed.
Before opening a matter, confirm: Is this investigation attorney-directed? If it is not — if HR is running it with legal in an advisory role, or if it was not initiated at the direction of counsel for the purpose of obtaining legal advice — the privilege analysis changes materially and this skill's default labeling may be misleading. Flag that question to the attorney before creating any log or file.
If there is any doubt about privilege applicability, the attorney should resolve it before investigation files are created. Improperly labeled materials can create problems in discovery if privilege is later challenged.
标注不产生特权。上述页眉反映了预期的保护范围，必须包含在内——但它本身并不构成特权。某一输出是否真正享有特权，取决于调查是否由律师主导、创建文档的目的以及后续的使用或披露方式。
开启matter前请确认：本次调查是否由律师主导？如果不是——例如由HR主导、法律仅提供咨询，或者并非为获取法律建议而由律师发起——特权分析将发生重大变化，本Skill的默认标注可能具有误导性。在创建任何日志或文件前，请将该问题告知律师。
如果对特权的适用性存在任何疑问，律师应在创建调查文件前予以解决。标注不当的材料在后续特权受到质疑时，可能会在证据开示环节引发问题。

Purpose

目的

Internal investigations fail in two ways: coverage gaps (sources that were never gathered) and synthesis gaps (evidence that was gathered but never connected). This skill handles both — it tracks what has and hasn't been gathered, processes document dumps to surface what matters without burying the attorney, and maintains a structured log that can be turned into a privileged memo at any point.
内部调查的失败分为两种情况：覆盖缺口（从未收集到相关来源）和整合缺口（已收集到证据但未建立关联）。本Skill可同时处理这两类问题——它跟踪已收集和未收集的内容，处理文档集以提炼关键信息而不增加律师负担，并维护结构化日志，可随时转换为特权备忘录。

Privilege note

特权说明

All files created by this skill carry the privilege marking above. See the notice at the top of this skill for the full caveat on what that marking does and does not do.
本Skill创建的所有文件均带有上述特权标注。有关该标注作用的完整说明，请参阅本Skill顶部的须知内容。

Load context

加载上下文

Read 
~/.claude/plugins/config/claude-for-legal/employment-legal/CLAUDE.md
→ escalation table, any investigation protocols noted.
读取
~/.claude/plugins/config/claude-for-legal/employment-legal/CLAUDE.md
中的升级流程表以及任何已记录的调查规程。

Mode 1: Open a new matter

模式1：开启新matter

Triggered by 
/employment-legal:investigation-open
or "open an investigation" or "start an investigation into".
由
/employment-legal:investigation-open
命令或“开启调查”“启动针对……的调查”触发。

Step 1 — Intake

步骤1 — 案件受理

Ask the following in a single block:
To open the investigation log I need a few things:
The matter
  • What is the allegation or concern in plain terms?
  • Who is the complainant (or what triggered this — complaint, tip, audit, manager observation)?
  • Who is the respondent or subject?
  • What is the approximate timeframe the alleged conduct occurred?
  • Is this attorney-directed? (If yes: work product protection applies. If no: flag privilege risk before proceeding.)
Investigation type (helps me suggest the right sources checklist)
  • HR: harassment / discrimination / retaliation
  • Financial misconduct: expense fraud / procurement irregularities / embezzlement
  • Executive misconduct: COI / undisclosed relationships / governance failures
  • Whistleblower: retaliation for protected activity
  • Other: describe briefly
Representation and employer status (surfaces parallel legal frameworks that change interview procedure)
  • Is the respondent, the complainant, or any anticipated witness represented by a union or covered by a collective bargaining agreement? (If yes, flag for Weingarten research — representational rights at investigatory interviews may apply and change the interview protocol.)
  • Is the company a public employer (government entity, public university, state or municipal agency) or otherwise acting under color of state law? (If yes, flag for Garrity research — compelled statements in public-sector investigations have special use-immunity consequences and change how interviews must be conducted and documented.)
If either flag fires, research the applicable rules (NLRA / state public-sector labor statutes for Weingarten; 5th Amendment and the Garrity line of cases, plus any state analogs) before conducting interviews. Cite primary sources. Verify currency. Do not interview until the protocol is adjusted.
一次性询问以下内容：
要开启调查日志，我需要以下信息：
Matter信息
  • 用直白语言描述指控或关注点是什么？
  • 投诉人是谁（或触发调查的原因是什么——投诉、举报、审计、经理观察）？
  • 被调查人或调查对象是谁？
  • 被控行为发生的大致时间范围是什么？
  • 本次调查是否由律师主导？（如果是：适用工作成果保护。如果不是：继续前请标记特权风险。）
调查类型（帮助我推荐合适的来源清单）
  • HR类：骚扰/歧视/报复
  • 财务不当行为：费用欺诈/采购违规/挪用公款
  • 高管不当行为：利益冲突/未披露关系/治理失效
  • 举报人：因受保护活动遭报复
  • 其他：简要描述
代理关系与雇主状态（涉及可能改变访谈流程的并行法律框架）
  • 被调查人、投诉人或任何预期证人是否由工会代理，或受集体谈判协议覆盖？（如果是：标记需进行Weingarten规则研究——调查性访谈中的代理权利可能适用，会改变访谈规程。）
  • 公司是否为公共雇主（政府实体、公立大学、州或市政机构），或以州法律名义行事？（如果是：标记需进行Garrity规则研究——公共部门调查中的强制陈述具有特殊使用豁免后果，会改变访谈的进行和记录方式。）
如果任一标记触发，请在开展访谈前研究适用规则（Weingarten规则适用NLRA/州公共部门劳工法规；Garrity规则适用第五修正案及相关判例，加上任何州级类似规则）。引用原始来源，验证时效性。在调整规程前不得开展访谈。

Step 2 — Create the matter directory and files

步骤2 — 创建matter目录和文件

Create the following files:
~/.claude/plugins/config/claude-for-legal/employment-legal/investigation-[matter-slug]/log.yaml
:
yaml
undefined
创建以下文件：
~/.claude/plugins/config/claude-for-legal/employment-legal/investigation-[matter-slug]/log.yaml
:
yaml
undefined

[WORK-PRODUCT HEADER — per plugin config ## Outputs — differs by role; see 
## Who's using this
]

[工作成果页眉——根据插件配置## Outputs——因角色而异；详见
## Who's using this
]

matter: "[matter name]" matter_slug: "[slug]" opened: "[ISO date]" attorney_directed: [true/false] allegation: "[plain-language summary]" complainant: "[name/role or anonymous]" respondent: "[name/role]" conduct_timeframe: "[approximate dates]" investigation_type: "[HR/financial/executive/whistleblower/other]" status: open last_updated: "[ISO date]"
issues:
  • "[Issue 1 — derived from allegation, e.g. 'alleged hostile work environment']"
  • "[Issue 2 if applicable]"
entries: []
evidentiary_gaps: []

`~/.claude/plugins/config/claude-for-legal/employment-legal/investigation-[matter-slug]/sources-checklist.yaml`:

Generated from the investigation type. See sources checklist templates below.

`~/.claude/plugins/config/claude-for-legal/employment-legal/investigation-[matter-slug]/documents-reviewed.yaml`:

```yaml
matter: "[matter名称]" matter_slug: "[slug]" opened: "[ISO日期]" attorney_directed: [true/false] allegation: "[直白语言摘要]" complainant: "[姓名/职位或匿名]" respondent: "[姓名/职位]" conduct_timeframe: "[大致日期]" investigation_type: "[HR/财务/高管/举报人/其他]" status: open last_updated: "[ISO日期]"
issues:
  • "[问题1——源自指控，例如'被控存在敌对工作环境']"
  • "[如有问题2]"
entries: []
evidentiary_gaps: []

`~/.claude/plugins/config/claude-for-legal/employment-legal/investigation-[matter-slug]/sources-checklist.yaml`:

根据调查类型生成。请参阅下方的来源清单模板。

`~/.claude/plugins/config/claude-for-legal/employment-legal/investigation-[matter-slug]/documents-reviewed.yaml`:

```yaml

[WORK-PRODUCT HEADER — per plugin config ## Outputs — differs by role; see 
## Who's using this
]

[工作成果页眉——根据插件配置## Outputs——因角色而异；详见
## Who's using this
]

matter: "[matter name]" total_reviewed: 0 total_surfaced: 0 last_updated: "[ISO date]" documents: []
undefined
matter: "[matter名称]" total_reviewed: 0 total_surfaced: 0 last_updated: "[ISO日期]" documents: []
undefined

Step 3 — Sources checklist

步骤3 — 来源清单

Generate the appropriate checklist based on investigation type. Present it to the attorney and ask: "Does this fit your matter? Let me know if any items are not applicable (I'll mark them N/A) or if there are additional sources specific to this situation."
HR investigation sources (harassment/discrimination/retaliation):
yaml
sources:
  - id: 1
    source: "Complainant interview"
    status: open
    notes: ""
  - id: 2
    source: "Respondent interview"
    status: open
    notes: ""
  - id: 3
    source: "Witness interviews — identify from complainant and respondent accounts"
    status: open
    notes: ""
  - id: 4
    source: "Email/messaging review — parties, relevant date range"
    status: open
    notes: ""
  - id: 5
    source: "HR records — respondent's performance history, prior complaints,
             prior discipline"
    status: open
    notes: ""
  - id: 6
    source: "Prior complaints — any prior complaints against respondent in
             HR system"
    status: open
    notes: ""
  - id: 7
    source: "Comparator data — how were similar situations handled"
    status: open
    notes: ""
  - id: 8
    source: "Relevant policies — harassment, code of conduct, reporting
             procedures (version in effect at time of alleged conduct)"
    status: open
    notes: ""
  - id: 9
    source: "Org chart and reporting relationships at time of alleged conduct"
    status: open
    notes: ""
  - id: 10
    source: "Calendar records — any meetings or events mentioned in accounts"
    status: open
    notes: ""
  - id: 11
    source: "Upjohn warning documentation — confirm interviews were preceded
             by Upjohn warnings and documented"
    status: open
    notes: ""
Financial misconduct sources:
yaml
sources:
  - id: 1
    source: "Expense reports — subject, relevant period"
    status: open
    notes: ""
  - id: 2
    source: "Approval records — who approved the expenses or transactions"
    status: open
    notes: ""
  - id: 3
    source: "Vendor/contractor records — contracts, invoices, payment records"
    status: open
    notes: ""
  - id: 4
    source: "Financial system records — AP, GL entries for relevant accounts"
    status: open
    notes: ""
  - id: 5
    source: "Email/messaging review — subject, approvers, counterparties"
    status: open
    notes: ""
  - id: 6
    source: "Subject interview"
    status: open
    notes: ""
  - id: 7
    source: "Approver interviews"
    status: open
    notes: ""
  - id: 8
    source: "Counterparty/vendor interviews (if accessible)"
    status: open
    notes: ""
  - id: 9
    source: "Audit logs — system access logs for relevant accounts/systems"
    status: open
    notes: ""
  - id: 10
    source: "Prior audits or reviews covering the relevant period"
    status: open
    notes: ""
  - id: 11
    source: "Upjohn warning documentation"
    status: open
    notes: ""
Executive misconduct sources:
yaml
sources:
  - id: 1
    source: "Subject interview"
    status: open
    notes: ""
  - id: 2
    source: "Board/compensation committee records — relevant resolutions,
             minutes, approvals"
    status: open
    notes: ""
  - id: 3
    source: "Employment agreement and any amendments"
    status: open
    notes: ""
  - id: 4
    source: "Equity records — grants, exercises, vesting"
    status: open
    notes: ""
  - id: 5
    source: "Expense reports and approval records"
    status: open
    notes: ""
  - id: 6
    source: "Email/messaging review — subject, relevant counterparties"
    status: open
    notes: ""
  - id: 7
    source: "Conflict of interest disclosures (or absence thereof)"
    status: open
    notes: ""
  - id: 8
    source: "Outside business activity records"
    status: open
    notes: ""
  - id: 9
    source: "Witness interviews — direct reports, peers, board members"
    status: open
    notes: ""
  - id: 10
    source: "Prior complaints or concerns raised about subject"
    status: open
    notes: ""
  - id: 11
    source: "Upjohn warning documentation"
    status: open
    notes: ""
Whistleblower sources:
yaml
sources:
  - id: 1
    source: "Complainant interview"
    status: open
    notes: ""
  - id: 2
    source: "Original complaint or tip — written form if exists"
    status: open
    notes: ""
  - id: 3
    source: "Records related to the underlying allegation (the thing
             complainant blew the whistle on)"
    status: open
    notes: ""
  - id: 4
    source: "Records related to any adverse action taken against complainant
             after the protected activity"
    status: open
    notes: ""
  - id: 5
    source: "Decision-maker interviews — who made the adverse action decision"
    status: open
    notes: ""
  - id: 6
    source: "Comparator data — treatment of similarly situated employees
             who did not engage in protected activity"
    status: open
    notes: ""
  - id: 7
    source: "Email/messaging review — decision-makers, relevant timeframe"
    status: open
    notes: ""
  - id: 8
    source: "Timing analysis — proximity of protected activity to adverse
             action"
    status: open
    notes: ""
  - id: 9
    source: "Respondent/decision-maker interviews"
    status: open
    notes: ""
  - id: 10
    source: "Upjohn warning documentation"
    status: open
    notes: ""
After presenting the checklist, write it to 
~/.claude/plugins/config/claude-for-legal/employment-legal/investigation-[slug]/sources-checklist.yaml
.
根据调查类型生成相应清单。将其呈现给律师并询问：“这份清单是否符合你的matter需求？如有不适用的项，请告知我（我会标记为N/A）；如有针对本次情况的额外来源，也请告知。”
HR调查来源（骚扰/歧视/报复）：
yaml
sources:
  - id: 1
    source: "投诉人访谈"
    status: open
    notes: ""
  - id: 2
    source: "被调查人访谈"
    status: open
    notes: ""
  - id: 3
    source: "证人访谈——从投诉人和被调查人的陈述中确定"
    status: open
    notes: ""
  - id: 4
    source: "邮件/消息审查——相关方、相关日期范围"
    status: open
    notes: ""
  - id: 5
    source: "HR记录——被调查人的绩效历史、过往投诉、过往处分"
    status: open
    notes: ""
  - id: 6
    source: "过往投诉——HR系统中针对被调查人的任何过往投诉"
    status: open
    notes: ""
  - id: 7
    source: "对比数据——类似情况的处理方式"
    status: open
    notes: ""
  - id: 8
    source: "相关政策——骚扰政策、行为准则、举报流程（被控行为发生时生效的版本）"
    status: open
    notes: ""
  - id: 9
    source: "被控行为发生时的组织结构图和汇报关系"
    status: open
    notes: ""
  - id: 10
    source: "日历记录——陈述中提及的任何会议或活动"
    status: open
    notes: ""
  - id: 11
    source: "Upjohn警告记录——确认访谈前已发出Upjohn警告并记录"
    status: open
    notes: ""
财务不当行为来源：
yaml
sources:
  - id: 1
    source: "费用报告——调查对象、相关时段"
    status: open
    notes: ""
  - id: 2
    source: "审批记录——谁批准了费用或交易"
    status: open
    notes: ""
  - id: 3
    source: "供应商/承包商记录——合同、发票、付款记录"
    status: open
    notes: ""
  - id: 4
    source: "财务系统记录——相关账户的应付账款、总账条目"
    status: open
    notes: ""
  - id: 5
    source: "邮件/消息审查——调查对象、审批人、交易对手"
    status: open
    notes: ""
  - id: 6
    source: "调查对象访谈"
    status: open
    notes: ""
  - id: 7
    source: "审批人访谈"
    status: open
    notes: ""
  - id: 8
    source: "交易对手/供应商访谈（如可接触）"
    status: open
    notes: ""
  - id: 9
    source: "审计日志——相关账户/系统的系统访问日志"
    status: open
    notes: ""
  - id: 10
    source: "涵盖相关时段的过往审计或审查"
    status: open
    notes: ""
  - id: 11
    source: "Upjohn警告记录"
    status: open
    notes: ""
高管不当行为来源：
yaml
sources:
  - id: 1
    source: "调查对象访谈"
    status: open
    notes: ""
  - id: 2
    source: "董事会/薪酬委员会记录——相关决议、会议纪要、审批"
    status: open
    notes: ""
  - id: 3
    source: "雇佣协议及任何修正案"
    status: open
    notes: ""
  - id: 4
    source: "股权记录——授予、行权、归属"
    status: open
    notes: ""
  - id: 5
    source: "费用报告及审批记录"
    status: open
    notes: ""
  - id: 6
    source: "邮件/消息审查——调查对象、相关交易对手"
    status: open
    notes: ""
  - id: 7
    source: "利益冲突披露（或未披露情况）"
    status: open
    notes: ""
  - id: 8
    source: "外部商业活动记录"
    status: open
    notes: ""
  - id: 9
    source: "证人访谈——直接下属、同事、董事会成员"
    status: open
    notes: ""
  - id: 10
    source: "针对调查对象的过往投诉或关注点"
    status: open
    notes: ""
  - id: 11
    source: "Upjohn警告记录"
    status: open
    notes: ""
举报人来源：
yaml
sources:
  - id: 1
    source: "投诉人访谈"
    status: open
    notes: ""
  - id: 2
    source: "原始投诉或举报——如有书面形式"
    status: open
    notes: ""
  - id: 3
    source: "与底层指控相关的记录（即投诉人举报的事项）"
    status: open
    notes: ""
  - id: 4
    source: "与投诉人从事受保护活动后遭受的任何不利行动相关的记录"
    status: open
    notes: ""
  - id: 5
    source: "决策者访谈——谁做出了不利行动的决定"
    status: open
    notes: ""
  - id: 6
    source: "对比数据——未从事受保护活动的类似情况员工的待遇"
    status: open
    notes: ""
  - id: 7
    source: "邮件/消息审查——决策者、相关时间范围"
    status: open
    notes: ""
  - id: 8
    source: "时间分析——受保护活动与不利行动的时间间隔"
    status: open
    notes: ""
  - id: 9
    source: "被调查人/决策者访谈"
    status: open
    notes: ""
  - id: 10
    source: "Upjohn警告记录"
    status: open
    notes: ""
呈现清单后，将其写入
~/.claude/plugins/config/claude-for-legal/employment-legal/investigation-[slug]/sources-checklist.yaml
。

Mode 2: Add data

模式2：添加数据

Triggered by 
/employment-legal:investigation-add
or "add to the [matter] investigation" or when the attorney pastes documents or interview notes.
由
/employment-legal:investigation-add
命令或“添加到[matter]调查”，或律师粘贴文档、访谈笔记时触发。

Step 1 — Identify the matter

步骤1 — 确定matter

If multiple investigation folders exist in 
~/.claude/plugins/config/claude-for-legal/employment-legal/
, ask which matter this data belongs to. If only one, proceed.
如果
~/.claude/plugins/config/claude-for-legal/employment-legal/
目录下存在多个调查文件夹，请询问该数据属于哪个matter。如果只有一个，则继续。

Step 2 — Identify the data type

步骤2 — 确定数据类型

Ask (if not clear from context):
  • Interview notes (whose interview?)
  • Document batch (emails, records, files)
  • Attorney notes or observations
  • Upjohn warning confirmation
询问（如果上下文未明确）：
  • 访谈笔记（谁的访谈？）
  • 文档批次（邮件、记录、文件）
  • 律师笔记或观察结果
  • Upjohn警告确认

Step 3 — Document pull criteria

步骤3 — 文档提取标准

For any document batch, apply the following pull criteria. A document is surfaced if it meets ANY of the following. The criteria are intentionally set to pull slightly aggressively — it is better to surface a false positive than to miss a significant item.
Pull criteria:
  1. Contains the name of any party to the investigation (complainant, respondent, witnesses named in prior log entries)
  2. Was authored or received by a party during the key conduct timeframe
  3. Contains keywords related to the allegation type (identified at intake and from prior log entries — update the keyword list as new terms emerge from accounts)
  4. Contains explicit or implicit admissions ("I shouldn't have," "I know how this looks," "don't put this in writing," "delete this")
  5. Contains language contradicting any account already in the log — flag the specific contradiction and the log entry it conflicts with
  6. Contains language that would be sensitive in litigation: discriminatory terms, threats, discussions of protected characteristics or activities, financial irregularities matching the allegation pattern
  7. Is a document type that has been mentioned in prior accounts but has not yet appeared in the document set (e.g., a meeting was mentioned in an interview but no calendar invite has been reviewed) → log as evidentiary gap, not a surfaced document
Disposition for every document reviewed:
  • surfaced
    : meets one or more pull criteria — added to log as a log entry
  • reviewed-nothing-significant
    : reviewed, does not meet pull criteria — logged in documents-reviewed.yaml with one-line description only
After processing a document batch, report:
Document review complete.
Reviewed: [N] documents
Surfaced: [N] as potentially significant
Logged as reviewed / nothing significant: [N]
New evidentiary gaps identified: [N]

Surfaced items:
[list with one-line description and which pull criterion triggered]
This report is the answer to "what about missed needles." The pull criteria are documented, the surface ratio is visible, and the attorney can review the full document log at any time. In Q&A mode, "I have not seen any document on [topic] in the [N] documents reviewed" is a meaningful statement only because every document reviewed is logged.
对于任何文档批次，应用以下提取标准。文档只要符合任一标准即被提炼出来。标准设置得较为宽松——提炼出假阳性内容总比遗漏重要信息要好。
提取标准：
  1. 包含调查相关方的姓名（投诉人、被调查人、先前日志条目中提及的证人）
  2. 由相关方在关键行为时间范围内撰写或接收
  3. 包含与指控类型相关的关键词（在受理阶段确定，并从先前日志条目中提取——随着新术语从陈述中出现，更新关键词列表）
  4. 包含明确或隐含的承认（“我不该这么做”“我知道这看起来怎么样”“别写下来”“删掉这个”）
  5. 包含与日志中已有陈述矛盾的内容——标记具体矛盾点及其对应的日志条目
  6. 包含在诉讼中会敏感的内容：歧视性用语、威胁、受保护特征或活动的讨论、与指控模式匹配的财务违规行为
  7. 是先前陈述中提及但尚未出现在文档集中的文档类型（例如，访谈中提到了某次会议但未审查日历邀请）→ 记录为证据缺口，而非提炼出的文档
每份已审查文档的处理结果：
  • surfaced
    ：符合一项或多项提取标准——作为日志条目添加到日志中
  • reviewed-nothing-significant
    ：已审查，不符合提取标准——仅在documents-reviewed.yaml中记录一行描述
处理完文档批次后，提交报告：
文档审查完成。
已审查：[N]份文档
提炼出潜在重要内容：[N]份
记录为已审查/无重要内容：[N]份
识别出新的证据缺口：[N]个

提炼出的内容：
[列出一行描述及触发的提取标准]
该报告用于回应“是否遗漏关键信息”的问题。提取标准已记录，提炼比例可见，律师可随时查看完整文档日志。在问答模式下，“在已审查的[N]份文档中，我未看到任何关于[主题]的内容”这一表述具有意义，因为每份已审查文档都已记录在案。

Step 4 — Write log entries

步骤4 — 写入日志条目

For each surfaced item, append to 
log.yaml
:
yaml
- entry_id: [auto-increment]
  entry_type: [interview / document / attorney-note / gap]
  date_of_event: "[date the event occurred — not when logged]"
  date_logged: "[ISO datetime]"
  source: "[witness name/role, or document filename/description]"
  source_type: [complainant / respondent / witness / document / attorney-note]
  issues: ["[which investigation issue(s) this entry relates to]"]
  significance: [high / medium / background]
  summary: "[what this entry adds to the record — 2-5 sentences]"
  quote: "[verbatim quote if significant — otherwise empty]"
  contradicts_entry: [entry_id or null]
  corroborates_entry: [entry_id or null]
  credibility_note: ""
  pull_criterion: "[which criterion triggered — for documents]"
  privilege: attorney-work-product
For evidentiary gaps:
yaml
- gap_id: [auto-increment]
  description: "[what document/source should exist but hasn't been found]"
  identified_from: "[which log entry or account raised this]"
  source_to_obtain: "[where to get it]"
  priority: [high / medium / low]
  status: open
对于每个提炼出的内容，追加到
log.yaml
中：
yaml
- entry_id: [自动递增]
  entry_type: [interview / document / attorney-note / gap]
  date_of_event: "[事件发生日期——而非记录日期]"
  date_logged: "[ISO日期时间]"
  source: "[证人姓名/职位，或文档文件名/描述]"
  source_type: [complainant / respondent / witness / document / attorney-note]
  issues: ["[本条目涉及的调查问题]"]
  significance: [high / medium / background]
  summary: "[本条目为记录新增的内容——2-5句话]"
  quote: "[如有重要内容则引用原文——否则为空]"
  contradicts_entry: [entry_id或null]
  corroborates_entry: [entry_id或null]
  credibility_note: ""
  pull_criterion: "[触发的标准——针对文档]"
  privilege: attorney-work-product
对于证据缺口：
yaml
- gap_id: [自动递增]
  description: "[应存在但未找到的文档/来源]"
  identified_from: "[提出该缺口的日志条目或陈述]"
  source_to_obtain: "[获取渠道]"
  priority: [high / medium / low]
  status: open

Step 5 — Update sources checklist

步骤5 — 更新来源清单

If the data added corresponds to a checklist item, ask the attorney if it should be marked complete or in-progress. Do not auto-mark complete — the attorney decides when a source is adequately covered.
如果添加的数据对应清单中的某一项，请询问律师是否应标记为已完成或进行中。不得自动标记为已完成——由律师决定某一来源是否已充分覆盖。

Mode 3: Query the log

模式3：查询日志

Triggered by 
/employment-legal:investigation-query
or any question phrased against the investigation (e.g., "what did [witness] say about", "what documents corroborate", "what do we still need", "what's the strongest evidence on each side").
Read the full log before answering. Answer types:
Factual query ("what did X say about Y"): Answer from the log entries, citing entry IDs. If the log contains nothing on the topic: "I have not seen any information on [topic] in this investigation log ([N] entries reviewed). This may be worth flagging as a gap."
Conflict query ("where do accounts conflict"): Surface all contradicts_entry links. For each conflict: state what the conflict is, which entries are in tension, and what (if any) documentary evidence bears on the conflict.
Coverage query ("what do we still need" / "what are our gaps"): Read sources-checklist.yaml and evidentiary_gaps in log.yaml. Report:
  • Checklist items still open
  • Evidentiary gaps logged
  • Any accounts that reference sources not yet gathered
Strength query ("what's the strongest evidence on each issue"): For each issue in the log, identify: the highest-significance log entries, any documentary corroboration, and any unresolved conflicts. Present issue by issue.
Upjohn query ("have we documented Upjohn warnings"): Check checklist item and any log entries tagged as Upjohn documentation. Flag if not yet completed.
由
/employment-legal:investigation-query
命令或任何针对调查的问题触发（例如，“[证人]关于……说了什么”“哪些文档可佐证”“我们还需要什么”“各方面的最强证据是什么”）。
回答前请通读完整日志。回答类型包括：
事实查询（“X关于Y说了什么”）： 根据日志条目回答，引用条目ID。如果日志中无相关内容：“在本次调查日志中（已审查[N]条条目），我未看到任何关于[主题]的信息。这可能值得标记为缺口。”
矛盾查询（“陈述存在哪些矛盾”）： 列出所有contradicts_entry链接。对于每个矛盾：说明矛盾点、存在冲突的条目，以及（如有）与矛盾相关的书面证据。
覆盖查询（“我们还需要什么”/“我们的缺口是什么”）： 读取sources-checklist.yaml和log.yaml中的evidentiary_gaps。报告：
  • 仍处于开启状态的清单项
  • 已记录的证据缺口
  • 任何提及尚未收集的来源的陈述
强度查询（“每个问题的最强证据是什么”）： 对于日志中的每个问题，确定：最高重要性的日志条目、任何书面佐证、任何未解决的矛盾。按问题逐一呈现。
Upjohn查询（“我们是否记录了Upjohn警告”）： 检查清单项和任何标记为Upjohn记录的日志条目。如果尚未完成则标记。

Mode 4: Draft or update the memo

模式4：起草或更新备忘录

Triggered by 
/employment-legal:investigation-memo
or "draft the memo" or "update the memo".
由
/employment-legal:investigation-memo
命令或“起草备忘录”“更新备忘录”触发。

If no memo exists — first draft

若备忘录不存在——首次起草

Read the full log. Do not draft until the following are complete (warn if not):
  • At least one entry for each open issue
  • Complainant and respondent entries present
  • Sources checklist reviewed (flag any high-priority open items)
Draft the memo in the following structure, following standard internal investigation memorandum practice:
markdown
[WORK-PRODUCT HEADER — per plugin config ## Outputs — differs by role; see `## Who's using this`]

---

**MEMORANDUM**

To: [Attorney to fill in]
From: [Attorney to fill in]
Date: [Date]
Re: Internal Investigation — [Matter name]
Status: PRELIMINARY DRAFT

---
通读完整日志。在完成以下事项前不得起草（如未完成则发出警告）：
  • 每个未解决的问题至少有一条条目
  • 存在投诉人和被调查人的条目
  • 已审查来源清单（标记任何高优先级未完成项）
按照标准内部调查备忘录惯例，以以下结构起草备忘录：
markdown
[工作成果页眉——根据插件配置## Outputs——因角色而异；详见`## Who's using this`]

---

**备忘录**

致：[由律师填写]
自：[由律师填写]
日期：[日期]
主题：内部调查——[Matter名称]
状态：初步草稿

---

Executive Summary

执行摘要

[2-3 paragraphs: allegation in plain terms, investigation scope and methodology summary, key findings in bullet form (Sustained / Not Sustained / Inconclusive), recommended actions. Written last but appears first.]
[2-3段：直白语言描述指控、调查范围和方法摘要、关键结论（已证实/未证实/不确定）、建议行动。最后撰写但置于开头。]

Background and Scope

背景与范围

Triggering event: [What initiated the investigation]
Allegations investigated: [Each issue from the log as a numbered allegation]
Out of scope: [Anything explicitly not investigated and why]
Investigation period: [Dates of conduct alleged] Investigation conducted: [Date opened] to [present or close date]
触发事件：[启动调查的原因]
调查的指控： [日志中的每个问题作为编号指控列出]
超出范围：[明确未调查的事项及原因]
被控行为时段：[被控行为发生的日期] 调查开展时段：[开启日期]至[当前或结束日期]

Methodology

方法

Interviews conducted:
WitnessRoleDateNotes
[Populated from log entries with source_type = interview]
Documents reviewed: [Summary of document categories reviewed, volume, date range. Full document log is maintained separately.]
Other sources: [Any other sources from checklist — policies, HR records, etc.]
Limitations: [Any sources requested but not obtained, any constraints]
已开展的访谈：
证人职位日期备注
[从source_type = interview的日志条目中填充]
已审查的文档： [已审查文档类别的摘要、数量、日期范围。完整文档日志单独维护。]
其他来源： [清单中的任何其他来源——政策、HR记录等]
局限性：[任何请求但未获取的来源、任何限制]

Factual Findings

事实结论

[Organized by issue — one section per allegation. Not by witness, not purely chronological.]
[按问题组织——每个指控设一个章节。不按证人或纯时间顺序。]

Issue 1: [Allegation]

问题1：[指控]

[Narrative of what the evidence shows on this issue. Cite log entry IDs inline in brackets. Where accounts conflict, present the conflict directly — do not smooth it over. Documentary evidence presented with quotes where significant.]
[关于该问题的证据说明。在文中引用日志条目ID（置于方括号中）。陈述存在矛盾时直接呈现——不得掩盖。重要的书面证据需引用原文。]

Issue 2: [Allegation]

问题2：[指控]

[Same structure]
[Continue for each issue]
[相同结构]
[每个问题依次呈现]

Credibility Assessment

可信度评估

[Standalone section. Address only witnesses whose credibility is determinative — i.e., where the finding on an issue depends on which account is credited.]
[独立章节。仅针对可信度具有决定性的证人——即某一问题的结论取决于采信哪一方陈述。]

[Witness name/role]

[证人姓名/职位]

Internal consistency: [Consistent / Inconsistent — note specifics] Corroboration: [What documentary or other evidence corroborates or undermines the account] Motive: [Any reason to credit or discount the account] Demeanor: [Attorney's observations if interviews were in person — leave blank if not applicable or not observed] Assessment: [Credit / Do not credit / Partially credit — with basis]
内部一致性：[一致/不一致——说明具体情况] 佐证：[哪些书面或其他证据佐证或削弱陈述] 动机：[采信或不采信陈述的任何理由] 举止：[如为面对面访谈，律师的观察结果——如不适用或未观察则留空] 评估：[采信/不采信/部分采信——说明依据]

Relevant Policies

相关政策

[Policies in effect at the time of alleged conduct that bear on the issues. Cite the version. Do not cite policies that were adopted after the conduct.]
[被控行为发生时生效的、与问题相关的政策。引用版本。不得引用行为发生后制定的政策。]

Conclusions

结论

IssueFindingBasis
[Issue 1]Sustained / Not Sustained / Inconclusive[One sentence]
[Issue 2]......
Findings are based on a preponderance of the evidence standard.
问题结论依据
[问题1]已证实/未证实/不确定[一句话说明]
[问题2]......
结论基于优势证据标准。

Recommendations

建议

[Organized by action type:]
Disciplinary action: [If any — state the basis, not just the outcome] Policy or process changes: [If any gap in policies contributed] Training: [If indicated] Further investigation: [Any threads not fully resolved] Monitoring: [Any follow-up needed]
[按行动类型组织：]
纪律处分：[如有——说明依据，而非仅结果] 政策或流程变更：[如有政策缺口导致问题] 培训：[如有必要] 进一步调查：[任何未完全解决的线索] 监控：[任何后续跟进需求]

Appendix A: Chronology of Events

附录A：事件时间线

[Auto-generated from log entries sorted by date_of_event, not date_logged. Format: Date | Summary | Source (Entry ID)]
[根据日志条目中的date_of_event（而非date_logged）自动生成并排序。格式：日期 | 摘要 | 来源（条目ID）]

Appendix B: Documents Reviewed

附录B：已审查文档

[Summary table from documents-reviewed.yaml]

Write the draft to `~/.claude/plugins/config/claude-for-legal/employment-legal/investigation-[slug]/memo.md`.
[从documents-reviewed.yaml生成的摘要表格]

将草稿写入`~/.claude/plugins/config/claude-for-legal/employment-legal/investigation-[slug]/memo.md`。

If memo already exists — update

若备忘录已存在——更新

Read the memo and the log. Identify log entries added since the memo was last drafted (compare date_logged against memo's last-updated date).
Report what has changed:
Since the last memo draft ([date]), the following has been added to the log:

[N] new entries
New issues: [any]
New conflicts: [any]
Resolved gaps: [any]

Sections that need updating:
  Factual findings: [which issues are affected]
  Credibility: [any new credibility-relevant entries]
  Conclusions: [any findings that should be revisited]
  Appendix A: [N] new chronology entries
Ask: "Want me to update the full memo, or just the affected sections?"
Apply updates. Preserve prior drafting. Mark changed sections with 
[UPDATED: date]
until the attorney reviews.
读取备忘录和日志。识别自备忘录上次起草以来新增的日志条目（比较date_logged与备忘录的最后更新日期）。
报告变更内容：
自上次备忘录草稿（[日期]）以来，日志中新增了以下内容：

[N]条新条目
新问题：[如有]
新矛盾：[如有]
已解决的缺口：[如有]

需要更新的章节：
  事实结论：[受影响的问题]
  可信度：[任何与可信度相关的新条目]
  结论：[任何需重新审视的结论]
  附录A：[N]条新时间线条目
询问：“需要我更新完整备忘录，还是仅更新受影响的章节？”
应用更新。保留先前起草内容。在律师审查前，对变更章节标记
[UPDATED: 日期]
。

Mode 5: Draft audience summary

模式5：起草受众摘要

Triggered by 
/employment-legal:investigation-summary
or "draft a summary for [audience]".
Ask: who is the audience and what decision or action does this summary support?
HR summary (for HR decision on disciplinary action):
  • What happened (factual summary, no legal analysis)
  • Finding on each allegation (Sustained/Not Sustained/Inconclusive)
  • Recommended action
  • What is NOT in this summary: privilege analysis, credibility methodology, legal exposure assessment, attorney mental impressions
  • Header: "Confidential — HR Use Only — Do Not Distribute"
  • Do not include entry IDs or document citations — those stay in the memo
Leadership/Board summary (for governance decision):
  • The allegation and scope in one paragraph
  • Key findings
  • Business impact / exposure (high level — no specific legal analysis)
  • What the company is doing about it
  • Header: "[WORK-PRODUCT HEADER — per plugin config ## Outputs — differs by role; see 
    ## Who's using this
    ]"
Outside counsel briefing (handing off for litigation or deeper review):
  • Full context including legal exposure analysis
  • Open evidentiary threads
  • Credibility issues that remain contested
  • Documents that would be most significant in litigation
  • Header: "[WORK-PRODUCT HEADER — per plugin config ## Outputs — differs by role; see 
    ## Who's using this
    ]"
由
/employment-legal:investigation-summary
命令或“为[受众]起草摘要”触发。
询问：受众是谁，该摘要支持什么决策或行动？
HR摘要（供HR做出纪律处分决策）：
  • 事件经过（事实摘要，无法律分析）
  • 每个指控的结论（已证实/未证实/不确定）
  • 建议行动
  • 本摘要不包含的内容：特权分析、可信度评估方法、法律风险评估、律师心理印象
  • 页眉：“保密——仅供HR使用——不得分发”
  • 不得包含条目ID或文档引用——这些内容保留在备忘录中
领导层/董事会摘要（供治理决策）：
  • 一段文字说明指控和范围
  • 关键结论
  • 业务影响/风险（高层级——无具体法律分析）
  • 公司的应对措施
  • 页眉：“[工作成果页眉——根据插件配置## Outputs——因角色而异；详见
    ## Who's using this
    ]”
外部律师简报（移交用于诉讼或深入审查）：
  • 完整上下文，包括法律风险分析
  • 未解决的证据线索
  • 仍存在争议的可信度问题
  • 诉讼中最重要的文档
  • 页眉：“[工作成果页眉——根据插件配置## Outputs——因角色而异；详见
    ## Who's using this
    ]”

Consequential-action gate (respond to a demand or complaint)

重大行动闸门（回应诉求或投诉）

Before producing a summary, memo, or content intended for an external response (EEOC/DFEH/state agency charge response, plaintiff's-counsel demand letter response, regulator response, or any formal complaint reply): Read 
## Who's using this
in 
~/.claude/plugins/config/claude-for-legal/employment-legal/CLAUDE.md
. If the Role is Non-lawyer:
Responding to a demand, charge, or complaint has legal consequences — positions taken here are admissions in later proceedings, waivers of defenses can be inadvertent, and privilege over the underlying investigation can be lost. Have you reviewed this response with an attorney? If yes, proceed. If no, here's a brief to bring to them:
  • The allegation, the forum, and the deadline
  • What the investigation surfaced (findings by allegation; documents reviewed; witnesses interviewed; Upjohn warnings given or not)
  • Any unresolved evidentiary threads or credibility contests
  • What the proposed response says and what it implicitly concedes
  • Open questions and what's unresolved
  • What could go wrong (privilege waiver, inconsistent factual statements, missed affirmative defense)
  • What to ask the attorney (is this the right theory; are we preserving defenses; should an outside firm take this over; what needs redaction or a privilege log)
If you need to find an attorney, solicitor, barrister, or other authorised legal professional: contact your professional regulator (state bar in the US, SRA/Bar Standards Board in England & Wales, Law Society in Scotland/NI/Ireland/Canada/Australia, or your jurisdiction's equivalent) for a referral service. Agency and demand-letter responses are a place where untrained replies regularly create more exposure than the underlying allegation did.
Do not produce an external-response draft past this gate without an explicit yes. Internal memos, HR summaries, and leadership briefings used only within the organization do not trip this gate (but the privilege-formation caveat at the top of this skill still applies).
在生成用于外部回应的摘要、备忘录或内容前（如EEOC/DFEH/州机构指控回应、原告律师诉求函回应、监管机构回应或任何正式投诉回复）：请阅读
~/.claude/plugins/config/claude-for-legal/employment-legal/CLAUDE.md
中的
## Who's using this
部分。如果角色为非律师：
回应诉求、指控或投诉会产生法律后果——此处采取的立场将在后续程序中成为自认，可能无意中放弃抗辩，且底层调查的特权可能丧失。你是否已与律师审查过该回应？如果是，请继续。如果不是，以下是提交给律师的简要说明：
  • 指控、受理机构和截止日期
  • 调查发现的内容（按指控列出的结论；已审查的文档；已访谈的证人；是否发出Upjohn警告）
  • 任何未解决的证据线索或可信度争议
  • 拟议回应的内容及其隐含承认的事项
  • 未解决的问题和不确定事项
  • 可能出现的问题（特权放弃、事实陈述不一致、遗漏积极抗辩）
  • 需向律师询问的问题（这是否为正确的理论；我们是否保留了抗辩；是否应由外部律所接手；哪些内容需要编辑或列入特权日志）
如果你需要寻找律师、事务律师、出庭律师或其他授权法律专业人士：请联系你的专业监管机构（美国州律师协会、英格兰及威尔士SRA/律师标准委员会、苏格兰/北爱尔兰/爱尔兰/加拿大/澳大利亚律师协会，或你所在司法管辖区的等效机构）获取推荐服务。机构诉求函回应是未经培训的回复通常会比底层指控造成更大风险的领域。
未获得明确同意前，不得通过此闸门生成外部回应草稿。仅在组织内部使用的内部备忘录、HR摘要和领导层简报不会触发此闸门（但本Skill顶部的特权形成须知仍适用）。

What this skill does NOT do

本Skill不执行的操作

  • Make disciplinary decisions — it supports the attorney's findings, not HR's action
  • Guarantee privilege — privilege depends on how the investigation is structured, not on how the memo is labeled
  • Process documents it cannot read — if files are in formats that cannot be parsed, flag them for manual review
  • Conduct interviews — it logs interview notes, it does not interview witnesses
  • Replace Upjohn warnings — it tracks whether they were given, it does not give them
  • 做出纪律处分决策——它支持律师的结论，而非HR的行动
  • 保证特权——特权取决于调查的结构，而非备忘录的标注
  • 处理无法读取的文档——如果文件格式无法解析，标记为需人工审查
  • 开展访谈——它记录访谈笔记，而非访谈证人
  • 替代Upjohn警告——它跟踪是否已发出警告，而非发出警告

Close with the next-steps decision tree

以下一步决策树收尾

End with the next-steps decision tree per CLAUDE.md 
## Outputs
. Customize the options to what this skill just produced — the five default branches (draft the X, escalate, get more facts, watch and wait, something else) are a starting point, not a lock-in. The tree is the output; the lawyer picks.
以CLAUDE.md中
## Outputs
部分的下一步决策树收尾。根据本Skill刚生成的内容自定义选项——五个默认分支（起草X、升级、获取更多事实、观望、其他）是起点，而非固定选项。决策树即为输出内容，由律师选择。