Agent Catalog & Usage Examples

Here are the specific usage instructions for the agents available in this repository.

Cross-Border Finance Agent (cross-border-finance)

A comprehensive agent for cross-border financial tasks, including exchange rate locking, compliance checks, and tax reporting.

• Exchange Rate Lock Engine (UI) Launch the Streamlit UI to analyze exchange rates and get lock recommendations.

  make exec AGENT=cross-border-finance ARGS="ui"

• Compliance Self-Check Tool (UI) Launch the Streamlit UI to upload transaction CSVs and generate compliance reports (PDF/Excel).

  make exec AGENT=cross-border-finance ARGS="compliance"

• ATO Tax Report Tool (UI) Launch the Streamlit UI to generate ATO-compliant tax declaration PDFs from transaction data.

  make exec AGENT=cross-border-finance ARGS="tax-report"

• Blockchain Log Evidence Module (UI) Launch the Streamlit UI to hash logs and upload evidence to Sepolia testnet.

  make exec AGENT=cross-border-finance ARGS="blockchain-log"

Model Verify Agent (model-verify)

A utility agent to verify model access and API keys.

• Verify Model Access

  # Check if OPENAI_API_KEY works for gpt-3.5-turbo
  make exec AGENT=model-verify ARGS="--key-name OPENAI_API_KEY --model gpt-3.5-turbo"# Check with a custom base URL (e.g., for local models or other providers)
  make exec AGENT=model-verify ARGS="--key-name OPENAI_API_KEY --model gpt-3.5-turbo --base-url https://api.openai.com/v1"

Tavily Search Agent (tavily-search)

An agent that performs web searches using the Tavily API.

• Perform a Search

  # Basic search
  make exec AGENT=tavily-search ARGS="search 'latest financial news in Australia'"# Search with topic and depth
  make exec AGENT=tavily-search ARGS="search 'RBA interest rate' --topic finance --search-depth advanced"

DeepQuant Agent (deepquant)

An agent wrapping the DeepQuant backtesting engine.

• Run Backtest

  make exec AGENT=deepquant ARGS="Backtest AAPL from 2023-01-01 to 2023-12-31 with 10000 usd and sma 20 and sma 50"

User Management Framework (framework)

The project includes a reusable Flask-based user management framework located in src/framework.

Database Migrations

We use flask-migrate (Alembic) to manage database schema changes.

1. Setup Environment Ensure your .env file has the correct DATABASE_URL. If not set, it defaults to a local SQLite finagent.db.

2. Initialize Migrations (First time only)

export FLASK_APP=src/run_server.py
flask db init

3. Generate Migration Script Run this whenever you modify src/framework/models.py:

export FLASK_APP=src/run_server.py
flask db migrate -m "Describe your changes"

4. Apply Migrations Apply the changes to the database:

export FLASK_APP=src/run_server.py
flask db upgrade

5. Quick Start (Development) To initialize the database without migrations (for quick testing):

python src/framework/init_db.py