Back to Details

policy-monitor

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

/policy-monitor

/policy-monitor

Sweep mode (no argument or 
--sweep
):
  1. Read 
    ~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
    → outputs folder path, AI policy document, last sweep date.
  2. Use the framework below. Scan outputs folder for files since last sweep.
  3. For each output: extract approved practices → diff against current policy commitments and use case registry.
  4. Classify gaps: REQUIRED (policy misrepresents current practice) vs ADVISABLE (policy silent).
  5. For each gap: quote current policy, describe gap, draft suggested language.
  6. Flag any use cases in outputs not yet added to the 
    ~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
    registry.
  7. Present results to the human. Only after acknowledgment, update 
    Last policy sweep
    and 
    gaps_found
    in 
    ~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
    .
Direct query mode (with description argument):
  1. Read 
    ~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
    → current policy commitments, use case registry, actual policy document.
  2. Parse proposed practice. Diff against policy: use case coverage, automation level, affected parties, disclosure, vendor data use, oversight.
  3. Output: covered / missing / conflicting + suggested language for each gap + registry entry if needed + timing recommendation.
Recurring runs: Set up a recurring reminder in your own scheduler to run 
/ai-governance-legal:policy-monitor
weekly. Scheduled execution requires a scheduled-tasks integration, which is not bundled with this plugin.
/ai-governance-legal:policy-monitor
/ai-governance-legal:policy-monitor "We want to use AI to automatically flag expense reports for review"
扫描模式（无参数或
--sweep
）：
  1. 读取
    ~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
    → 输出文件夹路径、AI政策文档、上次扫描日期。
  2. 使用下方框架。扫描输出文件夹中自上次扫描以来的文件。
  3. 针对每个输出：提取已批准的实践→与当前政策承诺和用例注册表进行对比。
  4. 分类差距：REQUIRED（政策与当前实践不符）vs ADVISABLE（政策未提及）。
  5. 针对每个差距：引用当前政策、描述差距、草拟建议措辞。
  6. 标记输出中任何尚未添加到
    ~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
    注册表的用例。
  7. 向用户展示结果。仅在得到确认后，更新
    ~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
    中的
    Last policy sweep
    和
    gaps_found
    。
直接查询模式（带描述参数）：
  1. 读取
    ~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
    → 当前政策承诺、用例注册表、实际政策文档。
  2. 解析拟议实践。与政策对比：用例覆盖范围、自动化程度、受影响方、披露要求、供应商数据使用情况、监督机制。
  3. 输出：已覆盖/缺失/冲突 + 每个差距的建议措辞 + 必要时的注册表条目 + 时间建议。
定期运行： 在您的调度器中设置定期提醒，每周运行
/ai-governance-legal:policy-monitor
。计划执行需要任务调度集成，此插件未捆绑该功能。
/ai-governance-legal:policy-monitor
/ai-governance-legal:policy-monitor "We want to use AI to automatically flag expense reports for review"

Purpose

用途

AI policies drift from practice faster than almost any other policy document — the field moves quickly, use cases multiply, and each approved AIA or triage result represents a new commitment the policy may not have caught up with. An AIA approves a new AI use case with a human-oversight condition. A vendor AI agreement permits data processing the policy doesn't mention. A triage result marks a new category of deployment as conditional with a disclosure requirement. The policy sits there unchanged.
This skill catches the drift — either by crawling the outputs folder weekly, or by answering the direct question: "we're about to start doing X, what does that mean for our AI policy?"
The output is always the same: here's the gap, here's the suggested language.
AI政策比几乎任何其他政策文档都更快地与实践脱节——该领域发展迅速，用例不断增多，每个已批准的AIA或分类结果都代表了一项政策可能尚未跟上的新承诺。AIA批准了一项带有人工监督条件的新AI用例。供应商AI协议允许政策未提及的数据处理。分类结果将一类新的部署标记为带有披露要求的条件性部署。而政策却保持不变。
此技能可发现这种偏差——要么通过每周爬取输出文件夹，要么通过回答直接问题：“我们即将开始做X，这对我们的AI政策意味着什么？”
输出内容始终一致：此处是差距，此处是建议措辞。

Load current state

加载当前状态

Read 
~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
:
  • ## AI policy commitments
    — commitments extracted from the published policy
  • ## Use case registry
    — approved, conditional, and never use cases
  • ## Outputs
    — outputs folder path, AI policy document location, last sweep date
If 
## Outputs
contains 
[PLACEHOLDER]
:
"Outputs aren't configured yet. I can still run a direct-query check — describe what you're planning to do and I'll diff it against your current AI policy. To enable the crawl sweep, run 
/ai-governance-legal:cold-start-interview
and provide the outputs folder path."
Read the actual AI or acceptable use policy document from the path in 
## Outputs
→ AI policy document. The commitments in 
~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
are a summary; the actual document is authoritative for suggesting edits.
读取
~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
：
  • ## AI policy commitments
    — 从已发布政策中提取的承诺
  • ## Use case registry
    — 已批准、条件性和禁止使用的用例
  • ## Outputs
    — 输出文件夹路径、AI政策文档位置、上次扫描日期
如果
## Outputs
包含
[PLACEHOLDER]
：
"输出尚未配置。我仍可运行直接查询检查——描述您计划做的事情，我会将其与您当前的AI政策进行对比。要启用爬取扫描，请运行
/ai-governance-legal:cold-start-interview
并提供输出文件夹路径。"
从
## Outputs
中的路径读取实际的AI或可接受使用政策文档→AI政策文档。
~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
中的承诺是摘要；实际文档是建议编辑内容的权威依据。

Mode detection

模式检测

Sweep mode: No argument, 
--sweep
, or triggered by schedule. → Scan the outputs folder. Diff all outputs since last sweep against current policy.
Direct query mode: User provides a description of a proposed new AI practice. → Diff that practice against current policy and use case registry. Suggest updates.
扫描模式：无参数、
--sweep
或由计划触发。 → 扫描输出文件夹。将自上次扫描以来的所有输出与当前政策进行对比。
直接查询模式：用户提供拟议新AI实践的描述。 → 将该实践与当前政策和用例注册表进行对比。建议更新内容。

Mode 1: Sweep

模式1：扫描

Determine scope

确定范围

Read 
## Outputs
→ Last policy sweep date. Scan for output files in the outputs folder dated after that date. If no date is recorded, scan all files and note: "First sweep — scanning all outputs."
If the outputs folder is empty or has no new files since the last sweep:
"No new outputs since [last sweep date]. AI policy appears current with recent practice. Next scheduled sweep: [date]."
Do not update 
Last policy sweep
or 
gaps_found
automatically. After the sweep results are presented, wait for the human to acknowledge them ("sweep acknowledged," "results reviewed," or equivalent). Only then update 
~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
:
  • Last policy sweep: [date of acknowledgment]
  • gaps_found: [N]
    (number of REQUIRED + ADVISABLE gaps found in that sweep)
Updating the stamp before acknowledgment would let an unreviewed sweep silently roll forward and suppress the next sweep's attention to the same gaps.
读取
## Outputs
→Last policy sweep日期。扫描输出文件夹中该日期之后的输出文件。如果未记录日期，则扫描所有文件并注明：“首次扫描——扫描所有输出。”
如果输出文件夹为空或自上次扫描以来没有新文件：
"自[上次扫描日期]以来无新输出。AI政策似乎与近期实践保持一致。下次计划扫描：[日期]。"
请勿自动更新
Last policy sweep
或
gaps_found
。 在展示扫描结果后，等待用户确认（“扫描已确认”“结果已审核”或类似表述）。仅在此时更新
~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
：
  • Last policy sweep: [确认日期]
  • gaps_found: [N]
    （此次扫描发现的REQUIRED + ADVISABLE差距数量）
在确认前更新标记会让未审核的扫描自动推进，并抑制下次扫描对相同差距的关注。

What to read in each output type

每种输出类型需读取的内容

AIAs (AI Impact Assessments):
  • Extract: use case approved, AI system description, deployment mode (assistive / augmentative / automated), conditions imposed, affected parties, vendor used, any disclosure requirements to affected individuals
  • Flag: use cases not in the registry, use cases approved with conditions not reflected in policy, vendor added that policy doesn't cover, automated decision deployed where policy implies human oversight
Triage results (CONDITIONAL / APPROVED outcomes):
  • Extract: use case classified, tier assigned, conditions imposed
  • Flag: new use case categories not in registry, conditions that imply policy commitments (e.g., "must disclose to affected parties" — does the policy say you do this?), newly approved practices that expand policy scope
Vendor AI reviews (signed / approved):
  • Extract: vendor added, data use terms agreed to, any AI-specific provisions accepted that differ from standard positions in 
    ~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
  • Flag: vendors added whose data use terms the policy should reference (e.g., "we use third-party AI services and ensure they do not train on our data"), approved deviations from standard positions that the policy implies you hold
Use case registry updates:
  • If new entries were added to the registry since the last sweep (directly, not through an AIA), check whether the policy reflects those approved categories.
AIAs（AI影响评估）：
  • 提取：已批准的用例、AI系统描述、部署模式（辅助/增强/自动化）、施加的条件、受影响方、使用的供应商、对受影响个人的任何披露要求
  • 标记：注册表中未有的用例、带有政策未体现条件的已批准用例、政策未涵盖的新增供应商、在政策要求人工监督的场景中部署的自动化决策
分类结果（CONDITIONAL / APPROVED结果）：
  • 提取：分类的用例、分配的层级、施加的条件
  • 标记：注册表中未有的新用例类别、暗示政策承诺的条件（例如，“必须向受影响方披露”——政策是否有相关规定？）、扩大政策范围的新批准实践
供应商AI评估（已签署/批准）：
  • 提取：新增的供应商、已同意的数据使用条款、与
    ~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
    中标准立场不同的任何已接受AI特定条款
  • 标记：政策应引用其数据使用条款的新增供应商（例如，“我们使用第三方AI服务并确保其不会使用我们的数据进行训练”）、与政策暗示立场不符的已批准偏差
用例注册表更新：
  • 如果自上次扫描以来注册表中新增了条目（直接添加，而非通过AIA），检查政策是否反映了这些已批准的类别。

Gap identification

差距识别

For each flagged item, assess:
REQUIRED update — the policy makes a commitment that an output contradicts, or an approved use case has no policy coverage and affects external parties. Not updating creates a material misrepresentation.
Example: AI policy says "we do not use AI in employment decisions." An AIA approved an AI-assisted hiring screening tool with human review required. Policy needs updating — even with human review, AI is now involved in employment decisions. "We do not use AI" is no longer accurate.
ADVISABLE update — policy is silent but not in conflict. The practice is defensible without updating, but cleaner with it. Important when the practice affects external parties or creates a reasonable expectation.
Example: Policy says "we use AI to improve our products and services." An AIA approved an AI feature for customer support drafts. Policy technically covers it but is vague. Advisable to be more specific so customers know what they're interacting with.
针对每个标记项，评估：
REQUIRED更新——政策做出的承诺与输出内容矛盾，或已批准的用例无政策覆盖且影响外部方。不更新会导致重大不实陈述。
示例：AI政策称“我们不在雇佣决策中使用AI”。某AIA批准了一项需人工审核的AI辅助招聘筛选工具。政策需要更新——即使有人工审核，AI现在已参与雇佣决策。“我们不使用AI”的表述不再准确。
ADVISABLE更新——政策未提及但无冲突。无需更新即可维持实践的合理性，但更新后会更规范。当实践影响外部方或产生合理预期时尤为重要。
示例：政策称“我们使用AI改进产品和服务”。某AIA批准了一项用于客户支持草稿的AI功能。政策在技术上涵盖了该场景，但表述模糊。建议更具体，以便客户了解他们所交互的内容。

Sweep output format

扫描输出格式

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

*This sweep is derived from AIAs, triage results, and vendor AI reviews that carry the plugin's privilege/confidentiality marking. The sweep inherits that status. Distribute deliberately — forwarding gap findings outside the privilege circle can waive privilege on the underlying assessments.*
markdown
[工作产品标题——根据插件配置## Outputs——因角色而异；参见`## Who's using this`]

*本次扫描基于带有插件保密/特权标记的AIAs、分类结果和供应商AI评估。扫描结果继承该状态。请谨慎分发——将差距发现转发至特权圈外可能会放弃基础评估的特权。*

AI Policy Monitor — Sweep Report

AI政策监控——扫描报告

Date: [date] Outputs scanned: [N files] | New since last sweep: [N files] Gaps found: [N] REQUIRED | [N] ADVISABLE
日期： [日期] 扫描的输出文件： [N个文件] | 自上次扫描以来新增： [N个文件] 发现的差距： [N] 项REQUIRED | [N] 项ADVISABLE

REQUIRED updates

REQUIRED更新

[Gap 1 short name]

[差距1简称]

Source: [filename / output type that triggered this] What's happening: [plain description of the new practice] Current policy: [quote the relevant section — or "No coverage"] Gap: [what's missing or inconsistent]
Suggested language:
Add to / update [section name]: "[Drafted policy text — specific, consistent with house style of the actual policy]"
[repeat for each REQUIRED gap]
来源： [触发此差距的文件名/输出类型] 当前实践： [新实践的简明描述] 当前政策： [引用相关章节——或“无覆盖”] 差距： [缺失或不一致的内容]
建议措辞：
添加至/更新[章节名称]： "[草拟的政策文本——具体，符合实际政策的内部风格]"
[重复每个REQUIRED差距]

ADVISABLE updates

ADVISABLE更新

[Gap name]

[差距名称]

Source: [filename] What's happening: [description] Current policy: [quote or "Silent"] Suggested language:
Add to / update [section]: "[Drafted text]"
来源： [文件名] 当前实践： [描述] 当前政策： [引用或“未提及”] 建议措辞：
添加至/更新[章节]： "[草拟文本]"

No action needed

无需操作

[List outputs scanned where no gaps were found]
[列出扫描后未发现差距的输出文件]

Use case registry sync

用例注册表同步

[Any use cases approved since the last sweep that aren't yet in the 
~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
registry — suggest registry entries to add]
[自上次扫描以来已批准但尚未添加到
~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
注册表的所有用例——建议添加的注册表条目]

Next steps

后续步骤

  • Review REQUIRED updates — decisions needed before the associated use cases go live (or immediately if already live)
  • Review ADVISABLE updates — lower urgency, address at next policy refresh
  • Add new use cases to registry (if any flagged above)
  • Next scheduled sweep: [date]

---
  • 审核REQUIRED更新——相关用例上线前（或如果已上线则立即）需做出决策
  • 审核ADVISABLE更新——优先级较低，可在下次政策更新时处理
  • 将新用例添加到注册表（如果有上述标记的用例）
  • 下次计划扫描：[日期]

---

Mode 2: Direct query

模式2：直接查询

Parse the proposed practice

解析拟议实践

Extract from the user's description:
  • What AI system or capability is being introduced?
  • What does it do — assistive, automated decisions, content generation?
  • Who does it affect — employees, customers, third parties?
  • Which vendor or model is involved?
  • Is there human review, or is it fully automated?
  • Are affected parties told the AI is involved?
  • Any data flowing to a vendor that wouldn't be expected?
If the description is vague, ask one clarifying question. Don't run a long intake — direct query mode should be fast.
从用户描述中提取：
  • 将引入何种AI系统或功能？
  • 它的用途是什么——辅助、自动化决策、内容生成？
  • 它会影响谁——员工、客户、第三方？
  • 涉及哪个供应商或模型？
  • 是否有人工审核，还是完全自动化？
  • 是否告知受影响方AI的参与情况？
  • 是否有流向供应商的意外数据？
如果描述模糊，提出一个澄清问题。不要进行冗长的收集——直接查询模式应快速完成。

Policy diff

政策对比

Check the proposed practice against the current policy and use case registry:
CheckCurrent policy / registryProposed practiceVerdict
Use case category[registry — approved / conditional / never / not present][new use case]🟢 Covered / 🟡 Gap / 🔴 Conflict
Scope of AI use[what policy says AI is used for][new use]
Automated decisions[policy position on automation][is this automated?]
Disclosure to affected parties[what policy commits to][what this requires]
Vendor data use[policy position on vendor AI][this vendor's terms]
Human oversight[policy statement if any][what's actually in place]
将拟议实践与当前政策和用例注册表进行对比：
检查项当前政策/注册表拟议实践结论
用例类别[注册表——已批准/条件性/禁止/未存在][新用例]🟢 已覆盖 / 🟡 差距 / 🔴 冲突
AI使用范围[政策规定的AI用途][新用途]
自动化决策[政策对自动化的立场][是否为自动化？]
向受影响方披露[政策承诺的内容][此实践的要求]
供应商数据使用[政策对供应商AI的立场][该供应商的条款]
人工监督[政策声明（如有）][实际设置的监督机制]

Direct query output format

直接查询输出格式

markdown
undefined
markdown
undefined

AI Policy Check: [Proposed practice in one line]

AI政策检查：[拟议实践的一句话描述]

Bottom line: [POLICY UPDATE REQUIRED / ADVISABLE / NO UPDATE NEEDED]
总结： [需更新政策/建议更新/无需更新]

What's covered

已覆盖内容

[Aspects of the proposed practice already addressed — brief, confirms no change needed]
[拟议实践中已被涵盖的方面——简明扼要，确认无需更改]

What's missing

缺失内容

[Gap 1]

[差距1]

Current policy: [quote or "Silent"] What's needed: [why this gap matters — legal, reputational, or expectation reason]
Suggested language:
Add to [section]: "[Drafted text]"
当前政策： [引用或“未提及”] 所需内容： [此差距的重要性——法律、声誉或预期方面的原因]
建议措辞：
添加至[章节]： "[草拟文本]"

[Gap 2]

[差距2]

[same format]
[相同格式]

What conflicts

冲突内容

[Conflict 1 — if any]

[冲突1——如有]

Current policy says: [quote] Proposed practice does: [what conflicts] Resolution: [which one needs to change — usually practice adjusts to match policy, or policy is updated to a defensible new position; never silently accept both]
当前政策规定： [引用] 拟议实践： [冲突内容] 解决方案： [哪一方需要更改——通常是实践调整以符合政策，或政策更新为合理的新立场；切勿同时接受两者]

Use case registry

用例注册表

[If this use case isn't in the registry: "Add to 
~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
→ Use case registry:"]
| [use case] | [Approved/Conditional] | [conditions] | — |
[如果此用例不在注册表中：“添加至
~/.claude/plugins/config/claude-for-legal/ai-governance-legal/CLAUDE.md
→ 用例注册表："]
| [用例] | [已批准/条件性] | [条件] | — |

Timing

时间安排

[REQUIRED: "Policy update should happen before this practice goes live — or immediately if it's already running." ADVISABLE: "Can proceed; update at next policy refresh."]

---
[REQUIRED：“政策更新应在该实践上线前完成——如果已上线则立即更新。” ADVISABLE：“可推进实践；在下次政策更新时完成更新。”]

---

Suggested language quality standards

建议措辞质量标准

AI policy language is unusually prone to becoming outdated — the field moves fast and vague language ages better than specific commitments. When drafting:
  • Match the voice and style of the existing policy (read the actual document)
  • Prefer durable language: "AI-assisted" rather than naming specific models that will change; "automated or AI-assisted decisions" rather than technical descriptions
  • Don't draft commitments the team can't keep — "we always have a human review AI outputs" is broken the moment one automated workflow ships
  • When a policy position is genuinely changing (not just extending), say so explicitly: "This update reflects that we now use AI in [new category] — the previous language did not cover this."
  • For disclosure language: draft it to be readable by the affected party (employee, customer), not just legally accurate
Always say which section to add to. If the right section doesn't exist, suggest creating it and draft the header.
AI政策语言特别容易过时——该领域发展迅速，模糊语言比具体承诺更耐用。草拟时：
  • 匹配现有政策的语气和风格（阅读实际文档）
  • 优先使用耐用语言：“AI辅助”而非命名会变化的特定模型；“自动化或AI辅助决策”而非技术描述
  • 不要草拟团队无法遵守的承诺——“我们始终由人工审核AI输出”在首个自动化工作流上线时就会失效
  • 当政策立场确实发生变化（不仅仅是扩展）时，明确说明：“此更新反映我们现在在[新类别]中使用AI——之前的语言未涵盖此场景。”
  • 对于披露语言：草拟时要让受影响方（员工、客户）能够理解，而不仅仅是法律准确
始终说明要添加到哪个章节。如果合适的章节不存在，建议创建该章节并草拟标题。

Schedule integration

调度集成

The weekly sweep is designed to run on a recurring cadence. Set up a recurring reminder in your own scheduler to run 
/ai-governance-legal:policy-monitor
weekly. Scheduled execution requires a scheduled-tasks integration, which is not bundled with this plugin.
After each sweep, the Last policy sweep and gaps_found fields in 
## Outputs
are updated only once the human has acknowledged the sweep results (see "Determine scope" above).
每周扫描旨在按定期周期运行。在您的调度器中设置定期提醒，每周运行
/ai-governance-legal:policy-monitor
。计划执行需要任务调度集成，此插件未捆绑该功能。
每次扫描后，仅在用户确认扫描结果后，才更新
## Outputs
中的Last policy sweep和gaps_found字段（参见上方“确定范围”部分）。

Close with the next-steps decision tree

以后续步骤决策树结束

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.
根据CLAUDE.md的
## Outputs
，以后续步骤决策树结束。根据此技能生成的内容自定义选项——五个默认分支（草拟X、升级、获取更多事实、观察等待、其他）是起点，而非固定选项。决策树是输出内容，由律师选择。

What this skill does not do

此技能不执行的操作

  • It doesn't update the policy itself — it drafts suggested language and flags decisions, but a human reviews and approves every change.
  • It doesn't catch incoming regulations — that's 
    reg-gap-analysis
    . This skill monitors internal practice drift, not external legal changes.
  • It doesn't enforce that outputs are saved — if AIAs and triage results aren't being saved to the configured folder, the sweep won't find them. Direct-query mode works without saved outputs.
  • It doesn't read email, Slack, or informal decisions — only structured outputs saved to the configured folder.
  • It doesn't update the use case registry automatically — it flags registry gaps and drafts entries for human review before adding.
  • 不直接更新政策——它草拟建议措辞并标记决策，但所有更改都需人工审核和批准。
  • 不跟踪新法规——这是
    reg-gap-analysis
    的功能。此技能监控内部实践偏差，而非外部法律变化。
  • 不强制要求保存输出——如果AIAs和分类结果未保存到配置的文件夹，扫描将无法发现它们。直接查询模式无需保存输出即可运行。
  • 不读取电子邮件、Slack或非正式决策——仅读取保存到配置文件夹的结构化输出。
  • 不自动更新用例注册表——它标记注册表差距并草拟条目，供人工审核后添加。",