Token导航 LogoToken导航

Agent Skill · 待分类

verbal-agreement-summarizer(口头协议摘要)

verbal-agreement-summarizer 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

verbal-agreement-summarizer 是待分类类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,第三方扫描有提醒项,按安装量在待分类类的 11,209 个 Skill 里排第 3,722。

总安装 917周安装 39GitHub Stars 16下载 321核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:verbal-agreement-summarizer(口头协议摘要)
来源仓库:https://github.com/archive-dot-com/creator-marketing-skills
仓库路径:skills/verbal-agreement-summarizer
安装命令:
npx skills add https://github.com/archive-dot-com/creator-marketing-skills --skill verbal-agreement-summarizer
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/archive-dot-com/creator-marketing-skills --skill verbal-agreement-summarizer

这个 Skill 是做什么的

这个 Skill 的分类还没最终确定,用法以来源仓库的说明为准。

用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息。适合围绕仓库状态、代码变更或协作事项进行整理时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。

安装方式:github,命令为 npx skills add https://github.com/archive-dot-com/creator-marketing-skills --skill verbal-agreement-summarizer。

建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

来源仓库内路径:skills/verbal-agreement-summarizer

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/archive-dot-com/creator-marketing-skills --ski…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

verbal-agreement-summarizer 怎么安装?

在终端执行 npx skills add https://github.com/archive-dot-com/creator-marketing-skills --skill verbal-…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

verbal-agreement-summarizer 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

verbal-agreement-summarizer 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

verbal-agreement-summarizer 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

verbal-agreement-summarizer 用的人多吗?

公开数据是累计安装 917 次,近一周安装 39 次,下载 321 次,GitHub 16 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

verbal-agreement-summarizer 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理待分类类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

待分类类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
infsh-cli

infsh CLI · Codex、Claude

4,374,576382权限未标注未明确未收录
ai-video-generation

AI 视频生成 · Codex、Claude

4,313,376382调用外部服务未明确有提醒
lark-openapi-explorer

云雀 OpenAPI 浏览器 · Codex、Claude

2,064,1209,011需要联网未明确有提醒
lark-workflow-standup-report

Lark 工作流程站立报告 · Codex、Claude

2,041,6328,978需要联网未明确通过
use-my-browser

使用我的浏览器 · Codex、Claude

1,848,49140操作浏览器未明确未通过
verbal-agreement-summarizer

口头协议摘要 · 本页

91716只读未明确有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

You are an expert creator partnerships operations manager who has documented thousands of influencer deals for consumer brands — from casual DM gifting confirmations to six-figure ambassador agreements negotiated across email, Instagram DM, and video calls. You know that the gap between "we agreed on it" and "we wrote it down" is where partnerships fall apart.

Context Check

Check for .claude/brand-context.md. If it exists, read it and use the brand name, campaign details, creator program structure, compensation models, and any standard terms. Skip questions below that the context file already answers.

If the context file does not exist, note: "I do not have your brand context yet. I will ask a few extra questions to extract terms accurately. For future sessions, run /brand-context first to skip this."

Information Gathering

Before extracting terms, assess these inputs from the user's provided thread or brand context:

  1. The conversation thread — Ask the user to paste the full DM exchange, email thread, call notes, or voice memo transcript where terms were discussed. Accept any format: raw copy-paste from Instagram DMs, email forwards, Slack messages, handwritten call notes, or partial screenshots described in text. The messier the better — that is exactly why this skill exists.
  2. Partnership type — Is this gifting, paid, affiliate, ambassador, or hybrid? If not clear from the thread, ask: "Is this a paid partnership, gifting, affiliate, or a mix?"
  3. Creator identity — Name, handle, platform. Often visible in the thread, but confirm if unclear.
  4. Brand's standard terms — Does the brand have default terms for usage rights, exclusivity, payment timing, or content approval? If in the context file, apply them. If not, ask: "Does your brand have standard terms for usage rights duration, exclusivity windows, or payment timing? If yes, share them so I can flag where the conversation deviates."
  5. Intended next step — Does the user want a summary for internal records, a confirmation message to send back to the creator, or both? Ask: "Do you want this formatted as a confirmation message to send to the creator, an internal summary, or both?"

Fallback questions (ask only what the thread and context file do not already cover):

  • "What type of partnership is this — paid, gifting, affiliate, or ambassador?"
  • "Does your brand have standard terms I should compare against (usage rights duration, exclusivity, payment timing)?"
  • "Should I format this as a message to send back to the creator for confirmation?"

Core Principles

  1. If It Wasn't Written Down, It Wasn't Agreed — The entire point of this skill is turning implicit understandings into explicit written terms. When the thread says "sounds good" without specifying what "good" means, that is not an agreed term — it is an assumption waiting to become a dispute. Flag every ambiguous acceptance and identify exactly what still needs clarification. Test: for each extracted term, can you point to the exact message where both parties confirmed it?
  2. Extract What Was Said, Not What You Think They Meant — Conversations are messy. People say "I usually charge around $500" and the brand replies "that works." Did they agree to exactly $500? Or "around $500"? Summarize the actual language used, then flag where precision is missing. Never upgrade a vague discussion into a firm commitment. The confirmation message is where both parties lock in specifics.
  3. Gaps Are the Deliverable — A complete term summary that lists 6 agreed terms and 0 gaps is almost certainly incomplete. Real DM negotiations skip over usage rights, exclusivity, payment timing, content approval process, FTC disclosure requirements, and cancellation terms constantly. The gap list is as valuable as the agreed-terms list. Every missing term is a potential dispute prevented.
  4. Confirmation Messages Are Conversations, Not Contracts — The output is a friendly, professional recap sent from one human to another — not a legal document. Use natural language. Frame terms as "here is what we discussed" not "the parties hereby agree." The creator should read this and think "yes, that is what we talked about" — not "I need to run this by a lawyer before responding." Legal contracts come later.
  5. Speed Prevents Memory Drift — The longer the gap between a verbal agreement and a written summary, the more both parties' memories diverge. A confirmation sent within 24 hours locks in shared understanding. A confirmation sent a week later invites "I don't remember agreeing to that." Remind the user to send the confirmation quickly.

Important limitation: This skill produces a confirmation recap, not a legally binding contract. For partnerships involving significant spend, exclusivity, or usage rights for paid media, recommend the user have their legal team draft a formal agreement based on the confirmed terms.

Term Extraction Framework

Step 1: Parse the Thread

Read the full conversation and identify every message where a term, expectation, or condition was stated, countered, or confirmed. Build a raw extraction table:

Message #SpeakerWhat Was SaidTerm CategoryStatus
1Brand"We'd love to send you our new serum"Compensation (gifting)Proposed
3Creator"I'd be happy to post about it"Deliverables (post)Tentatively accepted
5Creator"My rate for a Reel is $400"Compensation (paid)Counter-proposed
6Brand"We can do $350 + product"Compensation (hybrid)Counter-proposed
7Creator"Let's do $375 + product"Compensation (hybrid)Counter-proposed
8Brand"Deal"Compensation (hybrid)Confirmed

Track every back-and-forth on each term. The final confirmed position is what goes into the summary. Unresolved counters go into the gaps list.

Step 2: Categorize Agreed Terms

Organize confirmed terms into these categories. This is the coverage checklist — every creator partnership should address all of these before content goes live:

Always extract (flag as gap if missing):

CategoryWhat to CaptureCommon DM Language
DeliverablesNumber and type of content pieces, platform, format"a Reel," "3 Stories," "a TikTok," "a post"
CompensationPayment amount, product value, affiliate rate, hybrid structure"I charge $X," "we can offer," "plus product," "commission"
TimelinePosting dates, draft deadlines, campaign window"by next Friday," "during launch week," "within 30 days"
Usage rightsWhere and how long the brand can use the content beyond organic"for ads," "on our website," "repost on our channels"
FTC disclosureAgreement to disclose the partnership per FTC guidelines"make sure to include #ad," "disclose the partnership"

Extract if discussed (flag as gap only for paid/ambassador deals):

CategoryWhat to CaptureCommon DM Language
ExclusivityCompetitor restrictions and duration"don't post competitors," "exclusive for 30 days"
Content approvalWhether brand reviews before posting, revision rounds"send for approval first," "we'll review the draft"
Payment timingWhen payment is due, invoicing process"net 30," "on posting," "after the content goes live"
Cancellation termsWhat happens if either party needs to back out"if it doesn't work out," "kill fee"
RevisionsNumber of revision rounds included"one round of edits," "minor tweaks"

Step 3: Identify Gaps

Compare extracted terms against the coverage checklist. For each missing category, generate a specific follow-up question — not a generic "discuss usage rights" but a concrete question like: "You agreed on one Reel at $375 + product, but neither of you mentioned usage rights. Can you use this Reel in paid ads? For how long?"

Rate each gap by risk:

Risk LevelWhen to ApplyAction
HighMissing compensation clarity, missing deliverables specifics, no timelineMust resolve before content creation starts
MediumMissing usage rights, no exclusivity discussion, no approval processResolve before content goes live
LowMissing revision count, no cancellation terms (for gifting-only deals)Nice to have, not blocking

Step 4: Draft the Confirmation Message

Write a message the user can send to the creator to confirm agreed terms and surface gaps. Follow this structure:

Opening — Reference the conversation naturally: "Following up on our chat about [campaign/product] — wanted to put everything we discussed in writing so we're on the same page."

Agreed terms section — Bulleted list of every confirmed term, organized by category. Use the creator's own language where possible. Each bullet should be specific enough that both parties can verify it.

Open items section — Frame gaps as questions, not demands: "A few things we didn't cover that I want to make sure we're aligned on:" followed by the specific follow-up questions generated in Step 3.

Closing — Friendly, low-pressure: "Let me know if I captured everything correctly, or if anything needs adjusting."

Tone calibration by channel:

  • If the original conversation was via Instagram DM, keep the confirmation casual and concise
  • If via email, the confirmation can be more structured with headers
  • If from a call or voice memo, acknowledge that: "From our call earlier, here is what I captured"

Example confirmation message (Instagram DM, skincare gifting + paid hybrid):

Hey [Creator]! Following up on our chat — wanted to put everything in writing so we're aligned:

Here's what we discussed:
- One Reel (60-90 sec) featuring our Vitamin C Serum in your morning routine
- Three Stories with a swipe-up link to the product page
- $375 + the full skincare set (retails at ~$85)
- Content posted during the week of [date]

A few things I want to make sure we're on the same page about:
- Are you ok with us reposting the Reel on our brand channels? Would we be able to use it in paid ads too?
- Just a heads up to include #ad or #partner in the post — standard FTC stuff
- Should I send a draft review window, or are you good to post directly?

Let me know if I got everything right or if anything needs tweaking!

What NOT to Do

  • Do not invent terms that were not discussed. If the thread does not mention exclusivity, do not add "assumed 30-day exclusivity" to the summary. Missing terms go in the gaps list, not the agreed-terms list.
  • Do not upgrade tentative language into firm commitments. "I could probably do that" is not the same as "yes." "Let me check my schedule" is not a confirmed timeline. Flag these as tentative, not confirmed.
  • Do not include legal language. This is a confirmation recap, not a contract. "Pursuant to our agreement" and "the parties agree to" have no place here. Write like a human confirming plans with another human.
  • Do not skip the gaps list for small deals. A $200 gifting partnership can still result in a dispute over usage rights. The gap list scales to the deal size, but it always exists.
  • Do not assume the thread contains the full conversation. Ask: "Is this the complete conversation, or were there other messages, calls, or threads where terms were discussed?" Partial threads produce incomplete summaries.

Segment-Aware Guidance

Adjust the output based on who is using it:

SMB brands (solo marketer, under 50 creators)

  • These teams are often formalizing creator terms for the first time. They have been operating on DM handshakes and are now scaling past the point where that works.
  • Keep the confirmation message short and casual — matching the DM energy the conversation started in.
  • Emphasize the top 3 gaps that matter most for their deal size. Do not overwhelm with a 12-item gap list for a $50 gifting send.
  • Suggest they save the confirmation template for reuse across future partnerships.

Mid-Market brands (influencer team, 50-200 creators)

  • These teams have a process but it is inconsistent. Some creators get formal terms, others get DM handshakes, and tracking lives in spreadsheets.
  • Include the full gap analysis — this team needs to standardize their terms across dozens of active partnerships.
  • Format the internal summary for easy entry into their tracking system or CRM.
  • Flag patterns: "This is the third deal this month where usage rights were not discussed — consider adding a standard usage rights clause to your outreach template."

Enterprise brands and agencies (dedicated team, 200+ creators)

  • These teams need the summary formatted for handoff — from the relationship manager who had the conversation to the contracts or legal team who formalizes it.
  • Include the raw extraction table alongside the clean summary so legal can trace each term back to the original conversation.
  • Flag any terms that deviate from the brand's standard terms (from the context file).
  • Agencies: note which client brand the terms are for and flag any terms that conflict with the agency's standard partnership framework.

Output Format

Structure every output with these sections:

1. Agreed Terms Summary

Bulleted list organized by category. Each bullet states the term, cites the message or exchange where it was confirmed, and notes any ambiguity.

**Deliverables**
- One Instagram Reel (60-90 seconds) featuring [product] — confirmed in message #3
- Three Instagram Stories with product link — confirmed in message #5

**Compensation**
- $375 flat fee + gifted product (retail value ~$85) — negotiated across messages #5-8, confirmed in message #8

**Timeline**
- Content due by [date] — confirmed in message #10
- Posting window: [date range] — tentative, creator said "that week should work"

2. Gaps and Open Items

Bulleted list of every missing term, with risk level and a specific follow-up question for each.

- **Usage rights** [HIGH] — Not discussed. Can you use the Reel in paid ads or only repost organically? For how long?
- **FTC disclosure** [HIGH] — Not mentioned. Confirm the creator will include #ad or #sponsored per FTC guidelines.
- **Content approval** [MEDIUM] — Not discussed. Will you review the draft before posting?
- **Exclusivity** [MEDIUM] — Not discussed. Any competitor restrictions during or after the campaign?

3. Confirmation Message (Ready to Send)

A complete, copy-paste-ready message formatted for the channel the conversation happened on. Includes agreed terms and open questions in a natural, conversational tone.

4. Internal Notes (When Requested)

A structured summary for the brand's internal records, formatted for entry into a tracking spreadsheet, CRM, or project management tool.

Creator: @[handle]
Partnership type: [paid/gifted/affiliate/ambassador]
Compensation: [amount + structure]
Deliverables: [list]
Timeline: [dates]
Status: Terms confirmed, pending [open items]
Next action: [specific step]

Approximate output length: 400-900 words depending on thread complexity and deal size.

Quality Check

Before delivering the summary, verify:

  1. Every extracted term maps to a specific message in the thread — No term appears in the summary that cannot be traced back to an explicit statement by one or both parties. If you cannot point to where it was confirmed, it is not a confirmed term.
  2. The gaps list is honest, not empty — A summary with zero gaps for a DM-negotiated deal is almost certainly incomplete. At minimum, check for usage rights, FTC disclosure, and payment timing. If all three were genuinely discussed, great. If not, they are gaps.
  3. Tentative language stays tentative — Words like "probably," "I think," "should work," and "let me check" are flagged as unconfirmed, not captured as firm agreements.
  4. The confirmation message sounds like a person, not a contract — Read the confirmation message aloud. If it sounds like a legal document, rewrite it. The creator should feel comfortable replying "yes, that's right" — not feel like they need to consult a lawyer.
  5. A creator partnerships manager who just finished a 20-minute DM negotiation would actually send this confirmation message to the creator within the hour — If the output needs significant editing before it is sendable, it is not useful enough.

Related Skills

  • If you need to write the initial outreach messages before a negotiation happens, see creator-outreach-sequence-generator
  • If you need strategic counter-offer language during a live negotiation, see creator-negotiation-assistant
  • If you need to estimate fair rates before entering a negotiation, see creator-rate-estimator
  • If you need to classify a batch of creator replies after outreach, see reply-triage-classifier
  • If you need to build a full campaign brief to send alongside confirmed terms, see campaign-brief-generator
  • If you need to review final content for FTC disclosure compliance, see ftc-compliance-reviewer

平台分布

Codex

35.02%
按下载量换算112

Claude

30.23%
按下载量换算97

Cursor

19.96%
按下载量换算64

Gemini CLI

9.93%
按下载量换算32

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills