
FinePrint AI Legal Risk & Contract Audit MCP Server
io.github.RitualDev-Lab由 GitHub 账号 RitualDev-Lab 发布v0.1.1更新于 Nov 7, 2026
Zero-cost in-memory legal terms, API policy & NDA scanner for autonomous coding agents
发布方自己的服务:audit.domains - Domain Valuation
概览
本地内存中运行的合同风险扫描器,为编码代理评估法律合同、NDA 与租约风险并生成修订建议。
- 可以做什么
- 扫描合同文本或 PDF、DOCX、Markdown 文件中的法律风险条款
- 依据 60 余条启发式风险特征将合同评为 A 至 F 级
- 按赔偿、责任上限、知识产权归属、续约与管辖等类别列出问题
- 生成建议修订条款与高管摘要评分卡
- 以 Markdown 差异或 DOCX 修订模式导出红线稿
- 适用场景
- 适合在签署前审阅供应商条款、NDA、主服务协议或商业租约,也适合让编码代理检查 API 与供应商政策条款。希望先做一轮风险初筛、又不愿把文件上传到云端服务的团队可用。
- 运行要求
- 以本地 stdio 进程运行,通过 npm 包安装;README 还提到预编译二进制与 Python SDK。README 称使用 Gemini 1.5 Flash 进行法律推理与修订生成,这意味着需要 Gemini API 密钥,但清单未声明任何环境变量或认证方式。仅支持桌面端。
安装
在 SourceWeft 中
- 打开 控制台中的 FinePrint AI Legal Risk & Contract Audit MCP Server,将其添加到工作区。
- 为需要使用其工具的对话启用该服务。
Desktop only,通过 STDIO。 STDIO 服务会启动本地进程,因此需要 SourceWeft 桌面宿主。
其他 MCP 客户端
参照 仓库 中的启动说明。
README
📜 FinePrint AI
In-Memory Contract, Lease & NDA Scanner & Redline Generator
[GitHub Release] [CLI Tool] [Python SDK] [NPM Package] [License] [Zero VPS Cost]
FinePrint AI is an in-memory, privacy-preserving legal contract auditor and redline synthesizer. Equipped with 60+ heuristic legal risk signatures, an in-memory PDF AST parser, and Gemini 1.5 Flash, FinePrint scans Master Services Agreements (MSAs), NDAs, commercial leases, and vendor terms to grade contract risk from A (Acceptable) to F (Unacceptable) and generate attorney-grade redline amendments in seconds with zero cloud server overhead ($0 VPS).
Download Release • CLI Quickstart • Python SDK • Architecture • Risk Categories • Enterprise Features • Examples
⚡ The Problem: The Fine Print Trap
Reviewing commercial contracts is slow, error-prone, and expensive ($500–$1,200/hr for outside counsel):
- Uncapped Indemnification: Hidden clauses forcing vendors or customers to bear unlimited financial liability for third-party IP claims.
- Asymmetric Liability Caps: One-sided limitations where the vendor's liability is capped at $100 while the customer remains fully exposed.
- Auto-Renewal Traps: Multi-year automatic renewals requiring 90- to 120-day advance cancellation notice.
- Unfavorable Offshore Jurisdictions: Mandatory arbitration clauses buried in remote or offshore court systems.
- Privacy & Client Confidentiality: Uploading non-public corporate contracts or NDAs to generic cloud AI services violates confidentiality terms.
FinePrint AI performs all text extraction and AST clause chunking locally in-memory, ensuring your private legal agreements are never leaked or stored.
🏛️ Architecture
FinePrint operates on an air-gapped, zero-knowledge in-memory architecture:
🎯 60+ Legal Risk Categories
🚀 Quickstart
1. CLI Quickstart
Download the pre-compiled binary from Latest Releases or install via npm:
Run an automated audit on any contract or lease:
Export an executive scorecard:
2. Python SDK Quickstart (pip install fineprint)
Ideal for legal tech engineering, DocuSign webhooks, and contract data pipelines:
See examples/audit_contract.py for a complete runnable Python walkthrough.
3. Node.js / TypeScript SDK
🏢 Enterprise Features ($50K–$250K ACV Capabilities)
- 📜 Corporate Legal Playbook Customizer: Upload company-specific fallback terms and non-negotiables (e.g. Acme Corp mandates New York governing law, $5M cyber insurance, and strikes all customer indemnities). FinePrint scores incoming vendor contracts against your unique corporate playbook.
- 🗄️ Bulk Contract Repository Auditing (Legacy Contract Scanner): Batch scan thousands of legacy NDAs, MSAs, and commercial leases stored across Google Drive, Dropbox, or AWS S3 to uncover latent liabilities, expiring renewal dates, and uncapped indemnifications in bulk.
- ✍️ e-Signature Interceptor & Approval Workflows (DocuSign / PandaDoc Webhooks): Intercepts contracts before execution: if a contract receives an 'F' (Unacceptable) rating, the workflow automatically locks signature execution and routes the redline proposal to senior legal counsel for sign-off.
📊 Performance Benchmarks
- Analysis Latency: $< 1.2\text{ seconds}$ on 40-page commercial PDF agreements.
- Privacy: 100% in-memory processing; zero persistent document storage.
- Infrastructure Cost: $0.00 / month ($0 VPS, $0 server maintenance).
- Test Verification: 50 / 50 passing tests across Node.js, AST parsers, and Python 3.8+.
🤝 Community & Support
- Feature Requests & Bug Reports: Open an issue in our Public Issue Tracker.
- Enterprise Licensing: Contact
[email protected]for corporate playbook customization, DocuSign interceptors, and bulk legacy audits. - Core Development: Developed and maintained by RitualDev Studio.
来源:README.md,提交 d8e3612
工具
0版本历史
1- v0.1.1最新Nov 7, 2026
评分
已满足适用检查项的 40%
该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。
- 近 30 天内可连通25 分
请保持端点能够响应 MCP 握手;SourceWeft 每 30 天探测一次。
- 已声明认证方式10 分
请在注册中心条目中声明凭据请求头,或保持端点能够响应探测,以便判定其认证方式。
- 已列出工具15 分
请让服务器在无凭据时响应 tools/list,或发布可在沙箱中无凭据启动的软件包。
- 描述必需15 分
- 图标10 分
- 文档15 分
- 许可证10 分
请在仓库或注册中心条目中声明许可证。
其他 Audit 集成
audit.domains - Domain Valuation
domains.audit官方发布者
确定性的域名估值服务:提供公允价值区间、真实可比成交记录、可用性查询,以及经 DNS 验证的待售域名市场。
Revenue Agent — OpenAPI Audit & Change Detection
news.prh官方发布者
一个远程 MCP 端点,用于审计 OpenAPI 规范并检测 API 变更,每次分析通过 Base 上的 x402 收取 0.10 USDC。

Mcp Supply Audit
由 GitHub 账号 kimmawla 发布
把 MCP 服务器当作供应链产物来审计:注册表条目、PyPI/npm 包、客户端配置与工具定义,全部为只读静态检查。

Arquestra Audit
由 GitHub 账号 pablo-ramossanchez 发布
发布前审核数据文件:每个图表一盏灯。你的数据永远不会到达服务器。AI 翻译

PDF Accessibility Audit
由 GitHub 账号 Veduis 发布
本地优先的 PDF 无障碍审计,提供带引用的发现结果和离线 HTML 报告。AI 翻译