← 返回 Skills 市场
duclawbot

Claude

作者 Duclawbot · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ 安全检测通过
636
总下载
0
收藏
11
当前安装
1
版本数
在 OpenClaw 中安装
/install claude
功能描述
Optimize long-context reasoning for commercial, legal, and high-stakes documents. Built for users who need stronger logic, cleaner structure, and deeper anal...
使用说明 (SKILL.md)

Claude

Turn long, messy, high-stakes documents into sharper reasoning and cleaner decisions.

Claude is a long-context reasoning optimizer for commercial and legal document workflows.

Use this skill when you need to:

  • analyze long contracts, memos, proposals, or negotiations
  • improve reasoning quality across dense written material
  • surface contradictions, weak logic, or missing assumptions
  • compress large documents into decision-ready summaries
  • strengthen document structure before review or approval
  • compare multiple versions of a long-form document

This skill does NOT:

  • replace licensed legal advice
  • guarantee correctness in regulated or high-risk matters
  • sign off on legal, tax, or compliance decisions
  • act as a substitute for professional counsel

What This Skill Does

Claude helps:

  • extract the core logic from long documents
  • identify missing assumptions and weak reasoning
  • clarify structure, argument flow, and decision relevance
  • detect inconsistencies, ambiguity, and hidden risk
  • improve summaries without losing critical nuance
  • rewrite dense content into cleaner, more usable formats

Best Use Cases

  • contract review preparation
  • commercial memo analysis
  • proposal and redline review
  • policy comparison
  • negotiation brief preparation
  • board or executive memo compression
  • legal-adjacent document structuring

What to Provide

Useful input includes:

  • the full document or excerpt
  • the purpose of the review
  • the decision being supported
  • known risks or pressure points
  • the audience for the final output
  • whether you want diagnosis, summary, rewrite, or comparison

Standard Output Format

DOCUMENT ASSESSMENT
━━━━━━━━━━━━━━━━━━━━━━━━━━ Purpose: [What this document is trying to do]
Audience: [Who the output is for]
Decision relevance: [Why this matters]

CORE LOGIC
━━━━━━━━━━━━━━━━━━━━━━━━━━

  • [Main claim / obligation / commercial point]
  • [Supporting logic]
  • [Critical assumption]

RISKS / WEAK POINTS
━━━━━━━━━━━━━━━━━━━━━━━━━━ ⚠️ [Ambiguity]
⚠️ [Contradiction]
⚠️ [Missing assumption]
⚠️ [Commercial or legal risk signal]

STRUCTURE IMPROVEMENTS
━━━━━━━━━━━━━━━━━━━━━━━━━━

  1. [How to make reasoning clearer]
  2. [How to reduce ambiguity]
  3. [How to improve decision usefulness]

RECOMMENDED NEXT STEP
━━━━━━━━━━━━━━━━━━━━━━━━━━

  • [Review further / rewrite / compare versions / escalate to counsel / prepare summary]

Reasoning Principles

  • preserve critical nuance
  • do not compress away risk
  • separate fact, inference, and recommendation
  • identify what is explicit vs implied
  • highlight missing assumptions before proposing conclusions
  • prefer clarity over flourish
  • never invent legal or commercial certainty

Execution Protocol (for AI agents)

When user provides a long document, follow this sequence:

Step 1: Parse context

Extract:

  • document type
  • intended audience
  • decision being supported
  • major obligations, claims, or asks
  • commercial or legal sensitivity

Step 2: Identify logic structure

Map:

  • what the document is saying
  • why it matters
  • what assumptions it depends on
  • what could fail under scrutiny

Step 3: Detect weaknesses

Check for:

  • ambiguity
  • contradiction
  • undefined terms
  • missing scope boundaries
  • missing decision logic
  • hidden risk transfer
  • structural confusion

Step 4: Improve usefulness

Depending on request:

  • summarize
  • rewrite
  • compare
  • diagnose
  • convert into brief / memo / checklist

Step 5: Guardrails

If legal or commercial certainty cannot be established from the text:

  • say so clearly
  • mark uncertainty
  • do not fabricate confidence

Activation Rules (for AI agents)

Use this skill when the user asks about:

  • long documents
  • contract-like text
  • legal-adjacent review
  • memo analysis
  • commercial document reasoning
  • dense text summarization with nuance
  • comparing two long versions of a document

Do NOT use this skill when:

  • user only needs a casual short rewrite
  • user needs creative writing instead of reasoning
  • user needs formal legal advice or sign-off
  • user asks for certainty where the text does not support it

If context is ambiguous

Ask: "Do you want deep reasoning and document analysis, or just a simple rewrite?"


Works Well With

  • @ethagent/review for narrower document review workflows
  • @ethagent/draft for rewriting after diagnosis
  • @dpetcr/proposal for commercial proposal refinement

Boundaries

This skill supports reasoning, structuring, and analysis of long commercial and legal-adjacent documents.

It does not replace:

  • licensed legal advice
  • contract execution authority
  • procurement approval
  • tax or compliance judgment

Use outputs as analytical support, not formal sign-off.

安全使用建议
This skill appears internally consistent and low-risk, but you should still: 1) confirm the skill's provenance (source/homepage in skill.json differs from the top-level metadata and the source is 'unknown'); 2) avoid uploading highly sensitive or regulated documents until you verify how the platform handles data (logging, third-party calls, retention); 3) test with non-sensitive samples to confirm outputs and behavior; and 4) if you need formal legal advice, escalate to licensed counsel—the skill explicitly disclaims that role.
功能分析
Type: OpenClaw Skill Name: claude Version: 1.0.0 The skill bundle provides structured instructions for an AI agent to perform document analysis, reasoning, and summarization for commercial and legal-adjacent texts. It contains no executable code, external network calls, or instructions to access sensitive system data, and includes clear guardrails against providing legal advice or fabricating information in 'skill.md'.
能力评估
Purpose & Capability
The name, description, and declared capabilities all describe long-form document analysis (contracts, memos, comparisons). There are no unrelated environment variables, binaries, or config paths requested that would be inconsistent with this purpose.
Instruction Scope
SKILL.md instructs the agent to parse and analyze user-provided documents and to follow clear guardrails (do not provide legal certainty). It does not instruct reading unrelated files, accessing environment variables, or sending data to external endpoints.
Install Mechanism
This is an instruction-only skill with no install spec and no code files to write or execute. That minimizes on-disk installation risk.
Credentials
The skill declares no required environment variables, credentials, or config paths. The absence of requested secrets is proportionate to a document-analysis helper.
Persistence & Privilege
The skill does not request always:true or other elevated persistence. It is user-invocable and can be used by the agent per normal policy, which is expected for skills of this type.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install claude
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /claude 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
- Initial release of the Claude skill for long-context reasoning in commercial and legal documents. - Optimizes logic, structure, and analysis across contracts, memos, proposals, and negotiations. - Surfaces contradictions, weak logic, and missing assumptions; compresses documents into decision-ready summaries. - Standardized assessment output format and reasoning protocols for clarity and risk awareness. - Clear guide on when to use (or not use) the skill, with built-in guardrails against unfounded certainty. - Provides integration suggestions with related review and drafting workflows.
元数据
Slug claude
版本 1.0.0
许可证 MIT-0
累计安装 11
当前安装数 11
历史版本数 1
常见问题

Claude 是什么?

Optimize long-context reasoning for commercial, legal, and high-stakes documents. Built for users who need stronger logic, cleaner structure, and deeper anal... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 636 次。

如何安装 Claude?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install claude」即可一键安装,无需额外配置。

Claude 是免费的吗?

是的,Claude 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Claude 支持哪些平台?

Claude 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Claude?

由 Duclawbot(@duclawbot)开发并维护,当前版本 v1.0.0。

💬 留言讨论