Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问许可证需确认审计通过

contract-redline合同红线

Agent Skill

contract-redline 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。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。

总安装

449

周安装

18

GitHub Stars

56

下载量

145
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

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:contract-redline(合同红线)
来源仓库:https://github.com/vm0-ai/vm0-skills
仓库路径:skills/contract-redline
安装命令:
npx skills add https://github.com/vm0-ai/vm0-skills --skill contract-redline
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/vm0-ai/vm0-skills --skill contract-redline

简介

contract-redline 提供基于组织谈判手册的合同逐条审阅与修订建议。

  • 适合在 Codex、Claude、Cursor、Gemini CLI 中进行 SaaS、采购或服务协议的风险排查时使用。
  • 可输出带追踪标记的替换语言、风险评级及协商话术,辅助法务决策。
  • 需用户预先配置 playbook 文件或遵循主流商业惯例作为分析基准。
  • 强调仅作辅助参考,最终条款效力取决于签约方合意与法律管辖有效性。

SKILL.md

Pre-Review Setup

Organizational Playbook

Before any analysis begins, look for a negotiation playbook configured in the user's local settings. A playbook codifies the organization's preferred positions, tolerable bands, and hard limits for every major provision category.

When no playbook exists:

  • Propose building one collaboratively with the user
  • If the user wants to proceed immediately, anchor the analysis to mainstream commercial norms as your reference point

Contextual Framing

Every review must start by establishing three things:

  1. Agreement category -- Determine whether this is a SaaS subscription, professional services engagement, software license, channel partnership, procurement arrangement, or another structure. The category dictates which provisions carry the most weight.
  2. Client posture -- Establish whether the organization sits on the buying side, selling side, licensing side, or partnership side. Protective language that benefits one party harms the other.
  3. Holistic reading -- Read the entire document end-to-end before marking up any single provision. Provisions operate as an interconnected system. An aggressive indemnity clause may be counterbalanced by a strong liability cap elsewhere.

Provision-by-Provision Analysis

Liability Caps and Damage Exclusions

What to examine:

  • Total cap structure: fixed dollar figure, fee multiple, or absence of any ceiling
  • Symmetry of the cap between the parties
  • Exceptions carved out from the cap and which party they favor
  • Whether indirect, consequential, special, and punitive damages are waived
  • Mutuality of the damages waiver
  • Exceptions to the damages waiver
  • Cap measurement window: per-incident, annual, or lifetime aggregate

Typical problems:

  • Cap pegged to a small fraction of fees (for example, three months of spend on a modest-value deal)
  • One-sided exceptions that hollow out the cap for the drafter's benefit
  • Sweeping exception language like "any breach of this Agreement" that renders the cap meaningless
  • Asymmetric damages waiver leaving one party exposed to consequential loss claims

Indemnification Provisions

What to examine:

  • Reciprocity: does each side indemnify the other, or is it one-directional
  • Triggering events: IP infringement, data incidents, personal injury, warranty breaches
  • Relationship to the liability cap: subject to cap, partially capped, or unlimited
  • Procedural mechanics: timely notice, who controls the defense, settlement authority
  • Duty of the protected party to minimize harm
  • Survival period after the agreement ends

Typical problems:

  • One-directional IP indemnification when both parties contribute intellectual property
  • Catch-all "any breach" triggers that effectively eliminate the liability ceiling
  • No right for the indemnifying party to direct the legal defense
  • Open-ended survival with no time boundary

Intellectual Property Rights

What to examine:

  • Background IP ownership: each party must retain what they brought in
  • Foreground IP: who owns work product created during the engagement
  • Work-for-hire designations and whether their reach is proportionate
  • License grants: breadth, exclusivity, geographic scope, sublicense rights
  • Open source exposure
  • Feedback provisions that grant rights over suggestions or improvements

Typical problems:

  • Overbroad assignment language that could sweep in the customer's pre-existing assets
  • Work-for-hire clauses extending well beyond the specific deliverables
  • Perpetual, irrevocable feedback licenses with no practical limit
  • License scope that exceeds what the business relationship actually requires

Data Protection Provisions

What to examine:

  • Whether a Data Processing Agreement or Addendum is needed and present
  • Controller/processor role allocation
  • Sub-processor engagement rights and change-notification obligations
  • Breach reporting window (must enable the controller to satisfy the 72-hour GDPR deadline)
  • International transfer safeguards: Standard Contractual Clauses, adequacy findings, binding corporate rules
  • Data return or destruction duties upon contract end
  • Security standards and the controller's audit entitlements
  • Processing purpose restrictions

Typical problems:

  • Personal data in scope but no DPA attached
  • Unrestricted sub-processor authorization with no advance notice
  • Breach notification window that exceeds regulatory deadlines
  • No transfer protections for data crossing international borders
  • Vague or missing data deletion commitments

Duration, Renewal, and Exit

What to examine:

  • Length of the initial commitment and any renewal periods
  • Auto-renewal mechanics and the window for opting out
  • Convenience termination: availability, required notice, early exit penalties
  • Cause-based termination: what qualifies as cause, whether a cure window exists
  • Post-termination obligations: data handback, transition support, surviving provisions
  • Wind-down logistics and timeline

Typical problems:

  • Extended initial lock-in with no convenience exit
  • Auto-renewal paired with a narrow opt-out window (such as 30 days before an annual renewal)
  • Termination for cause with no opportunity to remedy the breach
  • Weak or nonexistent transition assistance language
  • Survival provisions that effectively perpetuate core obligations

Dispute Resolution and Governing Law

What to examine:

  • Applicable law and jurisdiction selection
  • Resolution pathway: courts, arbitration, mandatory mediation step
  • Litigation venue and personal jurisdiction
  • Arbitral institution, procedural rules, and seat (if arbitration applies)
  • Jury trial waiver
  • Class action waiver
  • Fee-shifting for the prevailing party

Typical problems:

  • Inconvenient or obscure venue selection
  • Compulsory arbitration under rules that advantage the drafter
  • Jury waiver without compensating procedural safeguards
  • No graduated escalation mechanism before formal proceedings

Deviation Rating System

GREEN -- Within Bounds

The provision matches or improves upon the organization's baseline position. Any variation is commercially sensible and does not meaningfully shift risk.

Illustrations:

  • Liability cap set at 18 months of fees when the baseline calls for 12 months (favorable to the buyer)
  • Mutual confidentiality term of 2 years against a 3-year baseline (shorter but reasonable)
  • Governing law in a reputable commercial jurisdiction near the preferred one

Response: Note for transparency. No negotiation warranted.

YELLOW -- Push Back

The provision sits outside the baseline but within a zone where negotiation is realistic. The position is seen in the market but is not the organization's preference. Warrants attention and discussion, though not immediate escalation.

Illustrations:

  • Liability cap at 6 months of fees against a 12-month baseline (below standard yet negotiable)
  • One-directional IP indemnification when the baseline is mutual (common but not preferred)
  • Auto-renewal opt-out of 60 days when the baseline is 90 days
  • Acceptable but non-preferred governing law jurisdiction

Response: Draft specific replacement language. Supply a fallback if the primary ask is refused. Estimate the business consequence of accepting the term as-is versus negotiating.

RED -- Escalate Immediately

The provision breaches the acceptable range, trips a defined escalation trigger, or introduces material exposure. Requires review by senior counsel, outside legal advisors, or a business decision-maker with sign-off authority.

Illustrations:

  • No liability cap at all, or no limitation of liability provision
  • Unilateral, uncapped, broadly-scoped indemnification
  • Assignment of the organization's background intellectual property
  • Personal data processing with no DPA offered
  • Unreasonable restrictive covenants or exclusivity demands
  • Hostile jurisdiction combined with mandatory arbitration

Response: Articulate the precise exposure. Offer market-standard replacement language. Quantify potential downside. Recommend the appropriate escalation path.

Crafting Effective Redlines

Principles for producing markup that advances the negotiation:

  1. Supply exact text -- Deliver language that can be inserted verbatim, not abstract guidance.
  2. Stay commercially reasonable -- Aggressive overreach slows deals. Be firm on critical protections and pragmatic elsewhere.
  3. Include professional rationale -- Attach a concise justification suitable for sharing with opposing counsel.
  4. Offer a Plan B -- For every YELLOW item, provide a secondary position in case the first request is declined.
  5. Rank by importance -- Signal which markups are essential and which are strategic asks.
  6. Match the relationship context -- Calibrate tone depending on whether the counterparty is a new vendor, a long-standing partner, or a commodity supplier.

Markup Format

Present each proposed change as follows:

**Provision**: [Section number and title]
**Existing text**: "[verbatim excerpt from the agreement]"
**Proposed replacement**: "[specific new language]"
**Justification**: [One to two sentences explaining the rationale, appropriate for external sharing]
**Importance**: [Essential / Strongly Preferred / Optional]
**Fallback**: [Alternative position if the primary request is declined]

Negotiation Prioritization

Organize all proposed markups into three tiers to guide negotiation strategy:

Tier 1 -- Non-Negotiable (Walk-Away Items)

Provisions where the organization cannot execute the agreement without resolution:

  • Absent or grossly inadequate liability protections
  • Missing data protection requirements for regulated information
  • IP terms that jeopardize core business assets
  • Clauses that conflict with the organization's regulatory obligations

Tier 2 -- High Priority (Strong Preferences)

Provisions that materially affect the risk profile but allow room for negotiation:

  • Liability cap adjustments within the acceptable band
  • Indemnification scope and reciprocity improvements
  • Flexibility around termination and exit rights
  • Audit, inspection, and compliance verification rights

Tier 3 -- Strategic Concessions (Trading Material)

Provisions that strengthen the position but can be yielded to secure more important wins:

  • Preferred governing law when the alternative is still acceptable
  • Notice period fine-tuning
  • Minor definitional refinements
  • Insurance documentation requirements

Negotiation approach: Open with Tier 1 demands. Offer Tier 3 concessions as currency to lock in Tier 2 outcomes. Never yield on Tier 1 without escalating to authorized decision-makers.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

36.29%
按下载量换算53

Claude

29.92%
按下载量换算43

Cursor

16.78%
按下载量换算24

Gemini CLI

8.14%
按下载量换算12

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

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。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills