π¦
Finance Coach MCP
No description available
0 installs
Trust: 30 β Low
Other
Ask AI about Finance Coach MCP
Powered by Claude Β· Grounded in docs
I know everything about Finance Coach MCP. Ask me about installation, configuration, usage, or troubleshooting.
0/500
Loading tools...
Reviews
Documentation
π° Finance Coach MCP
A personal finance coach app built with Python, Streamlit, OpenAI GPT-4o-mini, and the MCP (Model Context Protocol) SDK.
β¨ Features
- π¬ AI Chat Agent β talk naturally to log expenses, income, budgets, and get advice
- π Analytics Dashboard β 4 detailed pages with charts, trends, and breakdowns
- π― Budget Tracking β set limits, get alerts when you're close or over
- π Health Score β 0β100 financial health score based on your data
- π€ MCP Server β 18 tools exposed via the Model Context Protocol
- πΎ Local JSON Storage β simple file-based persistence, no database needed
π οΈ Tech Stack
| Layer | Technology |
|---|---|
| UI | Streamlit |
| AI | OpenAI GPT-4o-mini |
| Protocol | MCP SDK (FastMCP) |
| Storage | JSON files |
| Charts | Plotly |
| Language | Python 3.11+ |
π Project Structure
finance-coach-mcp/
βββ dashboard/
β βββ app.py β Main AI chat dashboard
β βββ pages/
β βββ 1_expenses.py β Expense analytics
β βββ 2_income.py β Income analytics
β βββ 3_budgets.py β Budget tracking
β βββ 4_advice.py β AI advice history
βββ data/ β JSON storage (auto-created)
βββ mcp_server/
β βββ server.py β FastMCP server (18 tools)
β βββ tools/
β βββ expenses.py
β βββ income.py
β βββ budgets.py
β βββ summary.py
β βββ context.py
β βββ advice.py
βββ utils/
β βββ storage.py β JSON read/write helpers
β βββ ai_client.py β OpenAI client wrapper
βββ main.py β Entry point
βββ requirements.txt
βββ .env.example
π Getting Started
1. Clone & install
git clone https://github.com/chai-sura/finance-coach-mcp
cd finance-coach-mcp
python -m venv fin_env
source fin_env/bin/activate # Windows: fin_env\Scripts\activate
pip install -r requirements.txt
2. Set up your API key
cp .env.example .env
# Edit .env and add your OpenAI API key
.env:
OPENAI_API_KEY=sk-...
3. Run
# Dashboard only (recommended)
python main.py --mode dashboard
# MCP server only
python main.py --mode mcp
# Both together
python main.py
Dashboard opens at http://localhost:8501
π§ MCP Tools
The MCP server exposes 18 tools:
| Category | Tools |
|---|---|
| Expenses | log_expense, get_expenses, get_expenses_by_category, delete_expense |
| Income | log_income, get_income, get_income_by_source, delete_income |
| Budgets | set_budget, get_budgets, get_budget_alerts |
| Summary | get_summary, get_monthly_trend |
| Context | view_context, get_financial_health_score |
| Advice | get_advice, get_advice_history, get_latest_advice |
π¬ Example Chat Commands
"I spent $50 on groceries"
"I earned $3000 salary"
"Set $300 monthly budget for dining"
"Show my financial summary"
"Give me financial advice"
"What's my health score?"
β οΈ Notes
- JSON storage is for local/personal use only β not production-ready
- OpenAI API costs are minimal (GPT-4o-mini, ~$0.01 per conversation)
- Data is stored in the
data/folder β add to.gitignoreif sensitive
π License
MIT
