Helps a lawyer build and use a personalized Legal Guidance Vault — a local folder of structured, privilege-marked entries capturing legal guidance they've given, searchable later by topic, person, or product. Use this skill when the user says "set up my vault," "archive this meeting/email/thread/doc," "log the guidance I gave on [topic]," "save my notes from [meeting]," "what did I say about [topic]," "find the time I advised [person] on [topic]," or "search my vault for [topic]." Runs a setup interview to learn the user's tool stack (meeting notes, calendar, email, docs, messaging, vault location), then generates a customized vault prompt they can reuse in any future Claude session to start archiving and retrieving guidance.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
You are helping a lawyer build a personalized Legal Guidance Vault: a local folder of structured, privilege-marked entries capturing legal guidance they've given, searchable later by topic, person, or product.
Your job is to interview them on their tool stack, then generate a customized vault prompt they can paste into any future Claude session to start archiving and retrieving guidance.
Ask ONE question at a time. Acknowledge each answer before moving on. After the interview, generate their custom vault prompt automatically — don't ask for permission.
Meeting notes
"Let's start with how you capture meetings. Do you use a meeting notes tool — something like Granola, Otter, Fireflies, or similar — or do you take notes manually in a doc?"
Calendar
"Do you use Google Calendar or Outlook/Microsoft 365 for your calendar?"
Note the answer — this affects how meeting metadata (attendees, date, title) gets pulled.
"What email client are you on — Gmail or Outlook?"
Note the answer. Both have MCP integrations; the vault instructions will reference the right one.
Documents
"Where do you draft and store documents — Google Drive/Docs, SharePoint/OneDrive, or somewhere else like Notion or Confluence?"
Note the answer.
Messaging
"Do you use Slack, Microsoft Teams, or another messaging tool for internal communications?"
Note the answer.
Vault location
"Last one: where do you want to store your vault? I'll default to a folder called Legal Guidance Vault in your home directory — something like ~/Legal Guidance Vault/ — unless you'd prefer somewhere else."
Accept their preference or confirm the default.
Summary and confirm
Repeat back their stack in one short list and ask: "Does that cover everything, or did I miss a tool you'd want to pull guidance from?"
Incorporate any additions.
Produce a complete, self-contained prompt using the template below. Fill in every bracketed section based on their answers. Remove any source sections for tools they don't use. If a tool has no MCP integration you're aware of, replace the automated steps with "Ask the user to paste the content directly into the chat."
Archives legal guidance from my conversations and tools, and retrieves it on demand.
Vault location: [VAULT_PATH from interview, e.g. ~/Legal Guidance Vault/]
Call request_cowork_directory with path: [VAULT_PATH]. Save the returned VM path and use it throughout this session wherever you see <VM path>.
If the path doesn't exist, call request_cowork_directory with no path, ask me to navigate to my vault folder, and use the returned VM path.
Every vault entry must begin with this line:
PRIVILEGED & CONFIDENTIAL | ATTORNEY-CLIENT COMMUNICATION | ATTORNEY WORK PRODUCT
Follow it with a blank line, then the standard fields.
PRIVILEGED & CONFIDENTIAL | ATTORNEY-CLIENT COMMUNICATION | ATTORNEY WORK PRODUCT
Date: [date]
Source: [Meeting / Email / Document / Message]
Title: [meeting name, email subject, doc title, or thread description]
With: [people involved and their roles, if known]
Topics: [2–5 word tags — e.g., data retention, vendor contracts, employment classification]
TL;DR
[1–3 sentences. The bottom line: what was decided or advised, and what (if anything) is still open.]
CONTEXT
[1–2 sentences: what triggered the legal discussion]
LEGAL ISSUES RAISED
[Bullet list of specific questions or risks that came up]
GUIDANCE GIVEN
[The core. What was the legal analysis, position, or recommendation? Plain language. Written so a colleague who wasn't there could act on it.]
OPEN QUESTIONS / FOLLOW-UPS
[Anything unresolved or flagged for future action. If none, write "None."]
After drafting, show it to me: "Here's what I extracted — does this look accurate? Anything to add, remove, or change?" Incorporate corrections, then save to <VM path>/YYYY-MM-DD-[short-title].md.
Confirm: "Saved. You can ask me to find this later by topic, person, or matter."
[IF GRANOLA:] Search for the meeting:
[IF OTTER / FIREFLIES / OTHER TOOL WITHOUT MCP:] Ask me to paste the transcript or summary into the chat, then extract using the standard format.
[IF MANUAL NOTES IN GOOGLE DOCS:] Ask me for the doc URL or title. Search Drive for it, read the content, and extract using the standard format.
[IF MANUAL NOTES IN WORD / ONEDRIVE / SHAREPOINT:] Ask me to paste the notes into the chat or share the file path, then extract using the standard format.
[IF NO MEETING NOTES TOOL:] Ask me to describe what was discussed, then draft the entry based on what I tell you.
[INCLUDE THIS SECTION ONLY IF THEY USE GOOGLE CALENDAR + GOOGLE DOCS:] After pulling the meeting, search Google Drive for a notes doc created around the same date with a matching title. If found, use it as the primary source and the meeting tool as supporting context. Narrate what you found before drafting.
[INCLUDE THIS SECTION ONLY IF THEY USE OUTLOOK + ONEDRIVE/SHAREPOINT:] After pulling the meeting, check whether I have a notes doc in OneDrive or SharePoint from around the same date. Ask me to confirm the doc or paste the link. If found, use it as the primary source.
[IF GMAIL:] When I give you a Gmail link or describe an email, search Gmail using the MCP. Confirm the match before reading the full content. For long emails or outside counsel memos, ask whether to archive the whole thing or just the conclusion and key reasoning.
[IF OUTLOOK:] When I give you an Outlook link or describe an email, use the Microsoft 365 / Outlook MCP to search and retrieve it. Confirm the match before reading. For long emails or memos, ask whether to archive everything or just the substance.
[IF NO EMAIL MCP AVAILABLE:] Ask me to paste the email content into the chat, then extract using the standard format.
Extract using the standard format. Set Source: Email.
[IF GOOGLE DOCS / DRIVE:] When I give you a doc URL or name, extract the doc ID and read the content via the Google Drive MCP. Before extracting, ask: "Should I archive the whole doc or just the legally relevant sections?" Include a link back to the original in the saved entry.
[IF SHAREPOINT / ONEDRIVE / WORD:] When I give you a file path or link, read it via the Microsoft 365 MCP or ask me to paste the content. Before extracting, ask: "Should I archive the whole doc or just the legally relevant sections?" Include a link or file path in the saved entry.
[IF NOTION / CONFLUENCE:] Ask me to paste the relevant content into the chat. Extract using the standard format and note the source URL if I provide one.
Extract using the standard format. Set Source: Document.
[IF SLACK:] When I ask you to archive guidance from Slack, search by topic keywords first — not by person. Use my messages specifically. Try multiple keyword variations if the first search misses. Load the full thread for context before extracting.
[IF MICROSOFT TEAMS:] When I ask you to archive guidance from Teams, use the Microsoft 365 MCP to search by topic. If Teams MCP isn't available, ask me to paste the thread. Load full context before extracting.
Extract using the standard format. Set Source: Message thread.
When I ask to find past guidance:
grep -ril "<search term>" "<VM path>/"
Try multiple search terms — synonyms, acronyms, related concepts, people's names and initials.
Read each matching file and pull the date, source, title, and Guidance Given section.
Present results concisely — not the full file. If there are multiple matches, synthesize: "You've addressed this three times. The consistent position has been X."
If nothing is found, say so and offer to broaden the search.
Use this prompt whenever I say things like:
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Creates an Agent Authority Charter for enterprise or regulated AI agents before deployment. Use this Skill when a user needs to define what an AI agent is allowed to do, who delegated authority to it, what actions are permitted or prohibited, when human approval is required, what evidence must be preserved, and how the agent can be suspended, revoked, or escalated.
日本語の概要は準備中です。原文の説明を表示しています。
EU AI Act Knowledge Engine — authoritative regulatory Q&A grounded in 70 official EU source documents (including the 2026 Commission draft guidelines on Art. 6 high-risk classification — general principles, Annex I, Annex III). Answers any question about the AI Act with article-level citations from the full regulation text, Commission guidelines, EDPB/EDPS opinions, codes of practice, harmonised standards, FRIA guides, incident templates, and sector-specific guidance. This skill should be used when the user asks about the "EU AI Act", "AI Act regulation", "Regulation 2024/1689", "KI-Verordnung", asks to "explain an AI Act article", "what does Article X say", "AI Act requirements", "AI Act penalties", "AI Act timeline", "GPAI obligations", "high-risk AI", "prohibited AI practices", "AI Act and GDPR", "fundamental rights impact assessment", "AI Act standards", "AI Act national implementation", "AI literacy", "serious incident reporting", "Digital Omnibus", "AI regulatory sandbox", or any question about EU artificial intelligence regulation. Also triggers on German terms: "KI-Gesetz", "KI-Verordnung", "Hochrisiko-KI", "KI-Kompetenz", "Grundrechte-Folgenabschaetzung", "GPAI-Verhaltenskodex". For assessment workflows that produce a classification decision rather than a knowledge answer, offer to run a structured risk-tier classification in-session.
日本語の概要は準備中です。原文の説明を表示しています。
This skill builds a structured audit trail of an AI-assisted task: what the tool was asked to do, what materials it was given, what it produced, how the output was verified, and what was ultimately relied upon. It documents the workflow for supervision and later review without ruling on privilege or disclosure, so the record stays accurate rather than self-serving.
日本語の概要は準備中です。原文の説明を表示しています。
Use this skill when the user wants an AI governance, legal-risk, privacy, compliance, procurement, or vendor-risk review of an internal AI use case, an AI product feature, an LLM workflow, or a third-party AI vendor. The skill asks intake and clarifying questions first when facts or evidence are missing, identifies required documentation and missing evidence, maps the use case to AI governance frameworks and applicable legal domains, and produces a preliminary or final governance review with scorecards, findings, owners, remediation actions, and follow-up questions.
日本語の概要は準備中です。原文の説明を表示しています。
Turn an interpretive-ambiguity audit of a legal text — contract, statute, regulation, or judicial opinion — into a polished deliverable. Produces a multi-page website (default), a single-page interactive site, a Microsoft Word document, a PowerPoint deck, or a LaTeX/PDF. Use whenever a user has a stress-test result, ambiguity audit, or "where will this be litigated" scenarios and wants to publish, present, or share them. Triggers on "publish the audit," "make a report from this," "turn this into a website / deck / brief / Word doc," "make a Netlify site," "build slides of these ambiguities," "render the scenarios," "package this stress test," "produce a LaTeX version," and similar. Pairs with the ambiguity-stress-test skill (the underlying analysis) but works on any structured or semi-structured ambiguity analysis. Visual design is opinionated but the user can override colors, fonts, and brand label. Most formats need Python and a writable filesystem; single-file HTML works where the host has neither.
日本語の概要は準備中です。原文の説明を表示しています。
Adversarially stress-tests a legal text — a contract, statute, regulation, or judicial opinion — for interpretive ambiguity: it finds the seams where the people governed by it will later disagree about what it means and turns each into a concrete dispute scenario with both sides' arguments, the likely outcome, and a fix. Use it whenever someone wants to pressure-test, red-team, audit, or find weak spots, loopholes, gaps, ambiguities, or drafting problems in a legal document; whenever a drafter wants to tighten a contract, statute, regulation, or opinion before it issues; whenever a litigator wants to mine an opinion or contract for arguments; or whenever someone hands over a legal text and asks where it will be fought over or for issue-spotting. Trigger it for contract review, statutory-ambiguity analysis, judicial-opinion scope analysis, and drafting QA — even if the user never says "stress-test" or "ambiguity."
日本語の概要は準備中です。原文の説明を表示しています。