tasty-agent
About This Server
A Model Context Protocol server for interacting with TastyTrade brokerage accounts. This server enables Large Language Models to monitor portfolios, analyze positions, and execute trades through the TastyTrade platform.
Server Information
This webpage provides information about tasty-agent, a Model Context Protocol (MCP) server designed to interface with TastyTrade brokerage accounts. It allows Large Language Models (LLMs) to monitor portfolios, analyze positions, and execute trades on the TastyTrade platform. The repository is public and licensed under the MIT License.
ā Key Points:
* tasty-agent is an MCP server for interacting with TastyTrade accounts.
* It allows LLMs to monitor portfolios, analyze positions, and execute trades.
* The project is open-source and uses the MIT license.
* It requires Python 3.12, uv package manager, and a TastyTrade account.
* Requires the Claude Desktop application for full functionality.
š Main Findings:
* The software uses a Model Context Protocol server to allow interaction with TastyTrade brokerage accounts.
* The server enables Large Language Models to monitor portfolios, analyze positions, and execute trades.
* The provided instructions detail installation, authentication, and local development testing.
* It provides market analysis tools such as 'getmetrics' and 'getprices'
* There is trade management functionality with schedule, list, and remove tools.
š Details:
* Installation: Requires Python 3.12 and uv package manager. Uses `uvx tasty-agent` to run.
* Authentication: Requires setting up TastyTrade credentials via the command line for secure storage in the system's keyring. Command: `uvx tasty-agent setup`
* Tools: Offers tools for Portfolio Management, Trade Management, and Market Analysis:
* Portfolio Management: `plotnlvhistory`, `getaccountbalances`, `getopenpositions`, `gettransactionhistory`.
* Trade Management: `scheduletrade`, `listscheduledtrades`, `removescheduled_trade`.
* Market Analysis: `getmetrics`, `getprices`.
* Usage with Claude Desktop: Requires adding a specific configuration to the `claudedesktopconfig.json` file. Scheduled trades execute only when Claude Desktop is running.
* Debugging: Uses MCP inspector and provides file paths for logs on macOS and Windows.
* Development: Includes instructions for local development testing with the MCP inspector and Claude Desktop.
* GitHub Repository: The repository includes a README.md, LICENSE, pyproject.toml, and uv.lock files.
šÆ Conclusion:
tasty-agent provides an interface between LLMs and TastyTrade, enabling automated portfolio management and trading strategies. The project offers instructions for setup, authentication, and development, with a caution that scheduled trades require Claude Desktop to be actively running.
Server Features
plot_nlv_history
Plots account net liquidating value history over time
get_account_balances
Get current account balances
get_open_positions
Get all currently open positions
get_transaction_history
Get transaction history
schedule_trade
Schedule a trade for execution
list_scheduled_trades
List all pending scheduled trades
remove_scheduled_trade
Remove a scheduled trade
get_metrics
Get market metrics for specified symbols
get_prices
Get current bid and ask prices
Provider Information
Ferdousbhai
cloud Provider