Official agent skill

Stakeholder Summary

by anthropics in anthropics/claude-for-legal

Translates a contract review into a summary the business stakeholder will actually read.

OfficialApache-2.0Auto-check passedLegal & Compliance

Install Stakeholder Summary

skills CLI
$ npx skills add anthropics/claude-for-legal --skill stakeholder-summary -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install anthropics/claude-for-legal stakeholder-summary --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/anthropics/claude-for-legal.git skills-src && mkdir -p .claude/skills && cp -r skills-src/commercial-legal/skills/stakeholder-summary .claude/skills/stakeholder-summary && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
stakeholder-summary
GitHub stars
9.6k
Used in
2 other repos
Token cost
~3.3k tokens
SKILL.md length
1,392 words
Files
1
Skills in repo
147
Repo updated
First seen
Licence
Apache-2.0

At a glance

Translates a contract review into a summary the business stakeholder will actually read.

  • Works in 3 steps: Count the escalation targets the review… → Count the escalations actually routed.… → Reconcile. If N approvers were named and…
  • User says summarize for the business
  • SKILL.md covers Matter context, Destination check, Purpose and Which side?, plus 6 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Stakeholder Summary is an agent skill from anthropics/claude-for-legal, published by the product's own GitHub organization. Translates a contract review into a summary the business stakeholder will actually read. Not a legal memo — a two-minute answer to "can I sign this and what do I need to know." Use when user says "summarize for the business", "write this up for [stakeholder]", "explain this to procurement", "non-legal summary", or when a review is done and needs to go to someone outside legal.

Its SKILL.md is about 3.3k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Legal & Compliance, covering Legal research, Vendor and procurement management and Contract review. The repository describes itself as: A suite of plugins for legal workflows. The licence is Apache-2.0.

When your agent uses it

  • User says summarize for the business
  • Write this up for [stakeholder]
  • Explain this to procurement
  • Non-legal summary

Example prompts

  • “can I sign this and what do I need to know.”
  • “summarize for the business”
  • “write this up for [stakeholder]”
  • “/stakeholder-summary”

Workflow steps

3 steps, taken from the first numbered list in SKILL.md.

  1. Count the escalation targets the review named. Look for the routing / escalation block at the end of the review, or for per-finding…
  2. Count the escalations actually routed. Read the review folder (or matter folder) for escalation-*.md drafts produced by escalation-flagger…
  3. Reconcile. If N approvers were named and M drafts exist, (N − M) escalations have not been routed.

What it can do on your machine

Read from SKILL.md and the folder at commit 4a6c651. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are markdown).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Stakeholder Summary loads about 3.3k tokens when it runs. Until then it costs about 100 tokens; SKILL.md has 1,392 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~100
When it runs · the whole SKILL.md, loaded when a task matches
~3.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from anthropics/claude-for-legal at commit 4a6c651, republished under its Apache-2.0 licence (© anthropics). 1,392 words, ~3,275 tokens.

Download SKILL.mdSave it as .claude/skills/stakeholder-summary/SKILL.md (or your agent's skills folder).
name
stakeholder-summary
description
Translates a contract review into a summary the business stakeholder will actually read. Not a legal memo — a two-minute answer to "can I sign this and what do I need to know." Use when user says "summarize for the business", "write this up for [stakeholder]", "explain this to procurement", "non-legal summary", or when a review is done and needs to go to someone outside legal.

Stakeholder Summary

Matter context

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 /commercial-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/commercial-legal/matters/<matter-slug>/. Never read another matter's files unless Cross-matter context is on.


Destination check

Before producing output, check where it's going. If the user has named a destination (a channel, a distribution list, a counterparty, "everyone"), ask whether it's inside the privilege circle. Public channels, company-wide lists, counterparty/opposing counsel, vendors, and clients (for work product) waive the protection. When the destination looks outside the circle, flag it and offer (a) the privileged version for legal only, (b) a sanitized version for the broader channel, or (c) both — don't silently apply a privileged header and then help paste it somewhere the header won't protect it. See the canonical ## Shared guardrails → Destination check in this plugin's CLAUDE.md.

Purpose

The business owner who asked for this contract doesn't want a legal memo. They want to know: can I sign it, what's the catch, and what do I need to do. This skill takes a completed review and turns it into that.

Which side?

The underlying review memo was run against either the sales-side or the purchasing-side playbook. Carry that framing through. A purchasing-side summary tells the business owner "here's what we're getting and what we agreed to give up"; a sales-side summary tells them "here's what we're selling and what we're on the hook for." Check which side the review was run on (it should be noted at the top of the review memo) and match the voice. If it's not obvious from the memo, ask the lawyer before summarizing.

Audience calibration

Read ~/.claude/plugins/config/claude-for-legal/commercial-legal/CLAUDE.md → ## House style → who reads stakeholder summaries, how long should they be. If not specified, default to: procurement or a department head, two paragraphs max, no legal terms of art.

Different audiences need different summaries:

AudienceCares aboutDoesn't care about
ProcurementPrice, renewal mechanics, approval routingLiability cap structure
Department head (budget owner)Can their team use it, what happens if it breaks, costIndemnity scope
FinanceTotal cost of ownership, renewal price risk, off-balance-sheet commitmentsGoverning law
Security / ITData handling, subprocessors, SOC 2, where data livesEverything else
Executive sponsorIs this going to embarrass us, is legal a blockerDetails

Ask who this is for if it's not obvious from context.

The summary

Length cap — enforced

The summary is:

  • One paragraph for the verdict and what this is (business terms, plain English)
  • One paragraph for the catch — the thing the stakeholder would be surprised by later if nobody told them now
  • A 2-3 item checklist for what the stakeholder actually needs to do (at most three items; if you want a fourth, the first three aren't tight enough)
  • A one-line close with approval timing

Under 200 words total. If you're writing more, you're including detail the stakeholder doesn't need — they have the memo for that. This is the quick read before the stakeholder hits reply.

If the close needs a third paragraph, fold it into the checklist instead. Don't let the close grow into a fourth block.

Scope of quote — discipline

When quoting a contract clause (in the summary, in the "catch" paragraph, or in the checklist), quote the full conditional sentence, not a truncated version. A clause that reads "Except as expressly provided in the Order Form, renewal of promotional or one-time priced subscriptions resets to list price" means something different from "renewal resets to list price" — the truncation drops the condition and misrepresents what the term does.

If a full conditional quote doesn't fit the summary's length cap, paraphrase rather than truncate. "For promotional pricing, renewal resets to list" is a fair paraphrase; "renewal resets to list" is not — it promotes the exception to the rule.

Format

Prepend the work-product header from ~/.claude/plugins/config/claude-for-legal/commercial-legal/CLAUDE.md ## Outputs (it differs by user role — see ## Who's using this).

markdown
[WORK-PRODUCT HEADER — per plugin config ## Outputs]
<!-- Remove the header above if forwarding outside the legal-privileged circle (e.g., to a business stakeholder, counterparty, or vendor). Confirm the correct marking for your jurisdiction and matter before forwarding. -->

**[Counterparty] [Agreement type]** — [READY TO SIGN | NEEDS CHANGES | BLOCKED]

[One paragraph: what this agreement does, in business terms. Not "Master Services
Agreement for the provision of cloud-based analytics" — "this is the contract
for the dashboard tool the marketing team wants."]

[One paragraph: what the stakeholder needs to know. The catch, if there is one.
The thing that will surprise them later if nobody tells them now. E.g., "Heads
up: this auto-renews every year and we have to cancel 60 days out. I've added
it to the tracker but you should know." Or: "Clean agreement, no surprises,
cleared to sign."]

<!-- Do not claim "I've added it to the tracker" unless `renewal-tracker` has
actually been run for this contract — see Verify tracker entries before
asserting them below. -->

**Verify tracker entries before asserting them.** Before the summary says "I've added it to the tracker" (or any equivalent — "it's in the tracker," "tracked," "set a reminder"), verify that `renewal-tracker` has been run for this contract. Check the outputs folder or the matter folder for a `renewal-tracker` output that names this counterparty / agreement. If there isn't one:

- Either run `renewal-tracker` for this contract first, then write the summary.
- Or write the summary without asserting the tracker entry, and include an action item: "Add to renewal tracker — not yet done."

Claiming a tracker entry exists when it does not is worse than omitting the reassurance. The stakeholder then trusts the reminder that will never fire. If the truthful statement is "tracked," the skill runs the tracker. If it's "you should add this to your calendar — I haven't logged it," say that.

**What you need to do:**
- [ ] [Action item, if any — "confirm the team is okay with data living in EU"
  or "nothing — I'll route for signature"]

**Approval:** [who's approving and expected timing]
What to translate
Legal findingBusiness translation
"Liability capped at 12 months fees""If they break something, the most we can recover is a year's worth of what we paid them."
"No termination for convenience""Once we sign, we're locked in for the full term — we can't just cancel if we stop using it."
"Auto-renewal with 60-day notice""This renews automatically every year. To cancel, we have to tell them two months before the renewal date."
"No IP indemnity""If someone sues us claiming this tool infringes their patent, the vendor isn't on the hook to defend us."
"Subprocessor list not disclosed""We don't know what other companies will have access to our data through them."
"Data deletion within 30 days of termination""When we cancel, they delete our data within a month. Export anything you need before then."
"SLA credits capped at 10% of monthly fee""If the service goes down, we get a small credit back. It won't cover the cost of the downtime to the business."
Show full SKILL.md (530 more words)Show less
What NOT to include
  • Section numbers
  • Defined terms in quotes
  • The word "indemnification" (say "they cover us if" / "we cover them if")
  • The word "notwithstanding"
  • Risk matrices with colored dots (unless this stakeholder has specifically asked for them before)
  • Caveats about how this isn't legal advice — the stakeholder knows who sent it

When the review found problems

If the review has 🔴 or 🟠 issues, the summary still needs to be two paragraphs — but the second paragraph is "here's what we're pushing back on and why."

markdown
[WORK-PRODUCT HEADER — per plugin config ## Outputs]
<!-- Remove the header above if forwarding outside the legal-privileged circle. -->

**[Counterparty] [Agreement type]** — NEEDS CHANGES

[What it is, one paragraph.]

We're going back to them on [N] things before this is ready. The main one:
[the critical issue in plain English — "they want the right to use our data
to improve their product, which means our competitors' instance gets smarter
from our data"]. We've asked them to strike it. [Realistic assessment: "They'll
probably agree" / "This might be a sticking point — will keep you posted."]

**What you need to do:**
- [ ] Nothing yet — I'll let you know when it's back from them.
  OR
- [ ] [Business decision they need to make: "If they won't budge on X, are you
  okay with Y, or do we walk?"]

Handoffs

From vendor-agreement-review / saas-msa-review: Those skills produce the full memo. This skill reads the memo and compresses it. Don't re-review the contract — read the review.

To the stakeholder: Via whatever channel ~/.claude/plugins/config/claude-for-legal/commercial-legal/CLAUDE.md says. If Slack, keep it under 150 words. If email, the format above is fine as-is.

Escalation-fan-out reconciliation

The upstream review is a one-to-many producer: it can name five escalation targets (Deputy GC, CISO, Privacy Officer, CFO, business owner) across different findings. escalation-flagger routes one finding at a time. Without a reconciliation step, the Deputy GC sees the memo and the other four approvers never do.

Before producing the summary, read the upstream review memo and tally escalations:

  1. Count the escalation targets the review named. Look for the routing / escalation block at the end of the review, or for per-finding "escalate to [X]" tags. De-dupe by approver name — a reviewer named for two findings counts once.
  2. Count the escalations actually routed. Read the review folder (or matter folder) for escalation-*.md drafts produced by escalation-flagger since the review was written. Each draft names one approver.
  3. Reconcile. If N approvers were named and M drafts exist, (N − M) escalations have not been routed.

Include a short reconciliation block in the summary — above the checklist, below the catch paragraph:

markdown
**Escalation status:** [M] of [N] escalation targets routed. The following have not been routed and require action:
- [Approver name] — [one line on the finding that named them]
- [Approver name] — [one line]

If all N have been routed:

markdown
**Escalation status:** [N] of [N] escalation targets routed.

If the upstream review surfaced no escalations, omit the block.

Do not omit a named approver from the reconciliation because the stakeholder wouldn't recognize the name. Business stakeholders often do not know who the Privacy Officer or CISO is. The reconciliation is internal-facing — it tells the lawyer sending the summary whether all the routing is done, not the stakeholder. If the stakeholder-facing summary needs to stay narrow, the reconciliation can live in a "routing status" footer or attached note — but it has to exist. A summary that implies routing is complete when it is not is worse than no summary.

Word-count carve-out. The escalation reconciliation block is exempt from the 200-word cap. Length-cap discipline on the summary body stays; the reconciliation is housekeeping, not narrative.

When no escalation-flagger drafts exist. If the upstream review named approvers and no drafts are in the folder, treat the count as M = 0. The reconciliation block lists all N as unrouted. That is the finding.

A note on tone

Stakeholders remember two things about legal: did it block me, and did it make sense. This skill is how legal makes sense. Write like you're explaining it to a smart colleague over coffee, not like you're writing a memo to file.

If the honest summary is "this is fine, sign it," say that. Don't pad a clean review into three paragraphs to look thorough.

© anthropics, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in commercial-legal/skills/stakeholder-summary of anthropics/claude-for-legal.

Open the folder on GitHubat commit 4a6c651

Used in 2 other repositories

We found 2 copies of this SKILL.md (exact, near-identical or edited) in other folders, from 2 other GitHub owners. This page covers the copy in anthropics/claude-for-legal, which our catalogue first saw on October 7, 2026.

Compare with similar skills

Stakeholder Summary next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Stakeholder Summary compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Stakeholder Summary this skillanthropics/claude-for-legal9.6k2 repos~3.3kAutomated safety check: PassApache-2.0
Msa Review Commercial PurchaseLegalQuants/lq-ai150—~5.7kAutomated safety check: PassApache-2.0
Contract Checklistrevfactory/harness-1001.3k—~1kAutomated safety check: PassApache-2.0
Finnish Legal Document Reviewakunikkola/claude-for-legal-finland109—~3.4kAutomated safety check: PassMIT
Clause Comparatorrohasnagpal/legal-ai-skills175—~2.7kAutomated safety check: PassMIT
Contract Drafterrohasnagpal/legal-ai-skills175—~3.6kAutomated safety check: PassMIT

Similar skills

  • A skill your agent uses when the user uploads or pastes a Master Purchase Agreement, Master Supply Agreement, Master Goods Agreement, or commercial purchase MSA covering the procurement of physical…

    150 GitHub stars~5.7k tokensUpdated yesterday
    Legal & ComplianceAuto-check passed
  • Contract Checklist

    revfactory/harness-100

    Procurement contract review checklist. An agent skill from revfactory/harness-100.

    1.3k GitHub stars~1k tokensUpdated 6 mo ago
    Legal & ComplianceAuto-check passed
  • Finnish Legal Document Review

    akunikkola/claude-for-legal-finland

    Checks a Finnish legal document, such as a contract, statement, appeal or decision, in eight stages, looking for errors, gaps and risks before it is signed or submitted.

    109 GitHub stars~3.4k tokensUpdated 3 mo ago
    Legal & ComplianceAuto-check passed
  • Clause Comparator

    rohasnagpal/legal-ai-skills

    Compares the same clause or provision across draft rounds, precedents, public agreements or a portfolio and reports what differs in wording and effect.

    175 GitHub stars~2.7k tokensUpdated 8 days ago
    Legal & ComplianceAuto-check passed
  • Contract Drafter

    rohasnagpal/legal-ai-skills

    Drafts a complete contract from a term sheet, negotiated heads or plain instructions — parties, recitals, definitions, operative clauses, schedules and boilerplate — in a specified posture and…

    175 GitHub stars~3.6k tokensUpdated 8 days ago
    Legal & ComplianceAuto-check passed
  • Negotiation Position Planner

    rohasnagpal/legal-ai-skills

    Builds an opening position, fallback, minimum acceptable position, walk-away line and concession logic for contract negotiations.

    175 GitHub stars~2.3k tokensUpdated 8 days ago
    Legal & ComplianceAuto-check passed

More from anthropics/claude-for-legal

All 147 skills in this repo
  • Legal Clinic Client Intake

    anthropics/claude-for-legal

    Official

    Structures a legal clinic client intake interview and produces a case summary with cross-area issue spotting, conflict flags and triage classification.

    9.6k GitHub starsUsed in 3 repos~3.2k tokens
    Auto-check passed
  • Supervisor Review Queue

    anthropics/claude-for-legal

    Official

    Holds student work in a queue for a legal clinic professor to approve, edit-then-approve or return before anything reaches clients or courts.

    9.6k GitHub starsUsed in 3 repos~1.1k tokens
    Auto-check passed
  • Tabular Document Review

    anthropics/claude-for-legal

    Official

    Builds a review grid with one row per document and one column per data point, each cell cited to a verbatim quote, built for M&A diligence and other batch reviews.

    9.6k GitHub starsUsed in 3 repos~4.3k tokens
    Auto-check passed
  • Product Launch Legal Review

    anthropics/claude-for-legal

    Official

    Runs a category-by-category legal review of a product launch from a PRD or tracker ticket, calibrated to your team's framework, and writes a review memo in house format.

    9.6k GitHub starsUsed in 2 repos~5k tokens
    Auto-check passed
  • Legal Skills Registry Browser

    anthropics/claude-for-legal

    Official

    Searches watched registries for community legal skills, shows matches with descriptions and offers the full SKILL.md before anything is installed.

    9.6k GitHub starsUsed in 2 repos~620 tokens
    Auto-check passed
  • Contract Renewal Tracker

    anthropics/claude-for-legal

    Official

    Shows which contracts renew soon and when notice must be sent by, working from a maintained renewal register, and warns about missed cancellation windows.

    9.6k GitHub starsUsed in 2 repos~3.1k tokens
    Auto-check passed

Questions about Stakeholder Summary

What does Stakeholder Summary do?

Translates a contract review into a summary the business stakeholder will actually read. Stakeholder Summary is an agent skill from anthropics/claude-for-legal, published by the product's own GitHub organization. Translates a contract review into a summary the business stakeholder will actually read.

When should I use Stakeholder Summary?

Stakeholder Summary fits situations like: user says summarize for the business; write this up for [stakeholder]; explain this to procurement; non-legal summary.

How do I install Stakeholder Summary in Claude Code?

Run `npx skills add anthropics/claude-for-legal --skill stakeholder-summary -a claude-code`. Or copy the skill folder (commercial-legal/skills/stakeholder-summary in anthropics/claude-for-legal) into .claude/skills/stakeholder-summary in your project. Claude Code loads it when a task matches its description.

How do I install Stakeholder Summary in Codex?

Run `npx skills add anthropics/claude-for-legal --skill stakeholder-summary -a codex`. Or copy the skill folder (commercial-legal/skills/stakeholder-summary in anthropics/claude-for-legal) into .agents/skills/stakeholder-summary in your project. Codex loads it when a task matches its description.

Can I use Stakeholder Summary in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add anthropics/claude-for-legal --skill stakeholder-summary -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/stakeholder-summary, .gemini/skills/stakeholder-summary, .github/skills/stakeholder-summary and .opencode/skills/stakeholder-summary in your project.

What does Stakeholder Summary need to run?

SKILL.md names no scripts, command-line tools or credentials: Stakeholder Summary is instructions for the agent only.

Does Stakeholder Summary access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Stakeholder Summary safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Stakeholder Summary use?

Stakeholder Summary is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Stakeholder Summary use?

About 3.3k tokens (SKILL.md is roughly 13k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Stakeholder Summary?

Skills that share tags, products or a category with Stakeholder Summary: Msa Review Commercial Purchase (LegalQuants/lq-ai, 150 stars), Contract Checklist (revfactory/harness-100, 1.3k stars), Finnish Legal Document Review (akunikkola/claude-for-legal-finland, 109 stars) and Clause Comparator (rohasnagpal/legal-ai-skills, 175 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Stakeholder Summary?

anthropics (a GitHub organization, an official publisher) maintains it in anthropics/claude-for-legal, which has 9,629 GitHub stars. The repository holds 147 skills in this directory. The repository was last updated on September 29, 2026.

Source: anthropics/claude-for-legal on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.