shelter

What This Skill Does

The Shelter AI skill acts as an intelligent financial coach that bridges the gap between raw bank data and actionable money management. By leveraging Plaid for secure, read-only access to your financial accounts, the skill provides a comprehensive dashboard of your economic health. It translates complex transaction history, recurring subscriptions, and income streams into plain language that helps you make informed decisions. Whether you are checking your safe-to-spend balance, predicting upcoming cash crunches, or seeking advice on debt management, the Shelter skill offers a structured way to interact with your personal finances.

Installation

To install this skill, run the following command in your terminal:

clawhub install openclaw/skills/skills/code-with-brian/shelter

Ensure that you have your SHELTER_API_KEY and SHELTER_API_URL environment variables set appropriately in your environment for the agent to authenticate successfully.

Use Cases

• Daily Spending Checks: Quickly verify how much you can safely spend today without impacting future bill payments.
• Subscription Audits: Automatically identify 'zombie' subscriptions that are draining your account without providing value.
• Financial Forecasting: Project your bank balance into the future to see if you are heading toward a cash crunch or a period of surplus.
• Affordability Analysis: Simulate large purchases to see how they impact your medium-term financial stability before you pull the trigger.
• Coaching: Receive personalized financial guidance tailored to your specific spending patterns and current stress level.

Example Prompts

1. "Can I afford to buy this $200 pair of shoes today without missing my rent payment next week?"
2. "How much runway do I have left at my current rate of spending?"
3. "Are there any unnecessary subscriptions I should cancel to save money this month?"

Tips & Limitations

• Efficiency: Always start with the cheapest endpoints like /v1/status or /v1/runway. Only utilize the /v1/ask endpoint for truly complex, nuanced queries to keep your API costs low.
• Data Integrity: Always check the confidence score returned by the API. If it is below 50, treat the insights as limited or speculative.
• Roleplay: Remember that you are an AI financial coach. Do not simply list raw JSON data; interpret it for the user with an empathetic, helpful tone.
• Privacy: As this skill accesses sensitive financial data, ensure your environment is secure and that you are not exposing your API keys in shared codebases.