Skip to main content
Glama

TreasuryOS

基于同一引擎构建的两个工具。选择适合你的入口。


面向企业主

了解你是否准备好申请商业贷款——免费,仅需 5 分钟。

上传你的银行对账单和信用报告。获取 0-100 分的准备度评分,了解哪些因素对你不利,学习如何改进,找出你有资格申请的贷款,并获取一份银行所需材料清单。

无需账户。无需费用。没有推销。

立即尝试 →

你将获得:

  • 基于银行评估标准的准备度评分

  • 用通俗易懂的语言说明哪些因素对你有帮助,哪些因素对你不利

  • 适合你情况的贷款类型(SBA 贷款、信用额度、设备融资等)

  • 在申请前需要优先解决的问题清单

  • 与银行经理会面时所需携带的材料清单

如何使用:

  1. 点击上方链接

  2. 点击 "See a Demo Score" 查看其工作原理

  3. 在侧边栏切换到 "Upload My Files"

  4. 上传你的银行对账单、信用报告或财务文档

  5. 查看你的评分和行动计划


Related MCP server: Sablier MCP Server

面向银行家与客户经理

为你完成准备工作的投资组合仪表板。

搜索客户、调取资料、查看整个客户群的融资准备情况。相同的评分引擎,专为客户经理工作流打造——一个搜索栏即可获取任何客户的全貌。

打开客户经理仪表板 →

你将获得:

  • 包含所有客户搜索功能的投资组合视图

  • 包含信用、现金、契约和准备度评分的客户资料

  • 9 种分析工具:现金头寸、预测、营运资金、信用评估、契约监控、外汇风险、闲置现金、支付优化

  • 与 Claude Desktop 集成的 MCP 服务器,用于人工智能驱动的分析

快速入门 (Claude Desktop):

git clone https://github.com/GPBK-STUY/treasury-mcp-os.git ~/Desktop/treasury-mcp-os
bash ~/Desktop/treasury-mcp-os/setup.sh

查看 QUICKSTART.md 获取完整设置说明。


构建方式

一个分析引擎驱动一切。三个应用程序服务于不同的受众。

应用程序

目标用户

在线链接

文件

SMB 应用

企业主

启动

smb_app.py

客户经理仪表板

银行客户经理

启动

rm_dashboard.py

完整仪表板

高级用户

启动

app.py

评分引擎

六大类别,每项评分 0-100:

类别

衡量指标

个人信用

FICO 分数、信用卡使用率、还款记录、查询记录

企业信用

Paydex 分数、经营年限、供应商付款速度、留置权

现金头寸

现金总额、账户数量、90 天预测

营运资金

流动比率、收款速度、现金周期、运营资金储备

现有债务

契约合规性、违约情况、预警

文档资料

已上传的关键文档类型

分析工具 (9)

Cash Position Aggregator    Credit Report Parser
Idle Cash Scanner           Credit Position Assessor
Cash Flow Forecaster        Covenant Monitor
Working Capital Analyzer    FX Exposure Scanner
Payment Optimizer

文件上传

支持 CSV、Excel (.xlsx)、Word (.docx) 和 PDF 格式。通过扫描列标题自动检测内容类型,并映射到正确的数据格式。

本地运行

git clone https://github.com/GPBK-STUY/treasury-mcp-os.git
cd treasury-mcp-os
pip install -r requirements.txt
streamlit run smb_app.py        # Business owner app
streamlit run rm_dashboard.py   # RM dashboard
streamlit run app.py            # Full dashboard

文件结构

treasury-mcp-os/
├── smb_app.py             ← Business owner app
├── rm_dashboard.py        ← RM portfolio dashboard
├── app.py                 ← Full dashboard (all 11 pages)
├── server.py              ← MCP server for Claude Desktop
├── tools/                 ← Shared analysis engine (9 tools)
├── sample_data/           ← Demo data (Apex Manufacturing Corp)
├── portfolio_data/        ← Multi-client data for RM dashboard
└── setup.sh               ← One-command Claude Desktop installer

关于

由 Grant Page 构建——拥有 5 年商业银行从业经验,致力于打造本应早就存在的工具。

许可证

MIT

Available Tools

9 tools
analyze_working_capitalC

Analyze working capital health. All values in same currency (USD).

Inputs: current_assets (required), current_liabilities (required), annual_revenue (required), accounts_receivable (required), +2 more.

assess_credit_positionA

Assess combined credit + cash flow lending readiness. Integrates credit reports with cash position and covenant data to produce an overall credit rating, risk factors, lending capacity estimate, and cross-sell opportunities. Args: annual_gross_income (combined guarantor income for DTI calc, 0 to skip).

Inputs: annual_gross_income.

forecast_cash_positionB

Forecast cash position using historical transaction patterns. Args: horizon_days (default 90, max 365).

Inputs: horizon_days.

get_cash_positionA

Get aggregated cash position across all bank accounts. Returns balances by currency and account type.

Inputs: no named parameters.

monitor_debt_covenantsA

Monitor debt covenant compliance across credit facilities. Flags breaches and warnings (within 10% of threshold).

Inputs: no named parameters.

optimize_payment_timingB

Find early-payment discounts that beat cost of capital. Args: available_cash budget (0 = show all opportunities).

Inputs: available_cash.

parse_credit_reportA

Parse personal and business credit reports already on file. Extracts FICO scores, Paydex, utilization, payment history, derogatories, and public records from credit data pulled during the application process.

Inputs: no named parameters.

scan_fx_exposureA

Scan for unhedged FX exposures in vendor payables. Calculates Value at Risk and recommends hedging strategies.

Inputs: no named parameters.

scan_idle_balancesC

Scan for idle cash that could earn higher yields. Args: operating_reserve_pct (default 0.20), target_yield_bps (default 450).

Inputs: operating_reserve_pct, target_yield_bps.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.6/5.0

    Scored across 9 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose targeting specific financial analysis tasks, such as cash position forecasting, credit assessment, covenant monitoring, and FX exposure scanning. There is no overlap in functionality; for example, analyze_working_capital focuses on liquidity health, while assess_credit_position integrates credit and cash flow data, making misselection unlikely.

    Naming Consistency5/5

    All tool names follow a consistent verb_noun pattern using snake_case, such as analyze_working_capital, forecast_cash_position, and monitor_debt_covenants. This uniformity makes the tool set predictable and easy to navigate, with no deviations in naming conventions.

    Tool Count5/5

    With 9 tools, the server is well-scoped for treasury and financial management, covering key areas like cash analysis, credit, risk, and optimization. Each tool earns its place by addressing a specific need without redundancy, fitting the typical range of 3-15 tools for such a domain.

    Completeness4/5

    The tool set provides comprehensive coverage for treasury operations, including monitoring, forecasting, and optimization tasks. Minor gaps exist, such as the lack of tools for direct transaction execution or detailed reporting generation, but agents can likely work around these with the available analysis and scanning tools.

    Maintenance

    ActivityMaintained
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      Not graded
      quality
      D
      maintenance
      A production-grade MCP server that provides financial ML tools including RAG search, anomaly detection, contract summarization, vendor graph analysis, and model drift monitoring using entirely free, open-source components.
      1
      -
    • A
      license
      Not graded
      quality
      C
      maintenance
      MCP server providing deterministic accounting tools for AI agents, including bank statement parsing, document classification, money math, and webhook verification.
      2
      Apache 2.0
    • A
      license
      Not graded
      quality
      D
      maintenance
      MCP server for k-prevention, enabling AI assistants to access decrypted financial data, run Monte Carlo simulations, optimize scenarios, and manage scenarios locally while preserving end-to-end encryption.
      MIT