AI Prompt Library
Jumpstart your AI conversations with ready-to-use prompts that help you get accurate, real-time insights from your business banking data.
Act as a cash flow analyst.
Analyze the past [Insert Number, e.g., 12] months of transactions to:
OUTPUT: Format as a table: Week | Risk Rating | Why
At the end, add one line noting this is based on historical patterns and scheduled items visible in the data, and unscheduled expenses or revenue changes will affect actual results. Then generate the PDF summarizing the content.
RULES: If any portion of the prompt is incomplete, please flag it immediately and prompt me back to ensure all portions of the prompt are filled out.
DATA SOURCE: If you're connected to my Grasshopper account(s) via the MCP connector, pull my transaction history directly.
Act as a payment processing consultant.
My account shows net deposits and fee debits from [Insert Processor, e.g., Stripe, Shopify, or Square] over the last [Insert Number, e.g., 90] days.
Present as: a summary table (Gross Sales / Total Fees / Net Deposited / Effective Rate / Target Rate / Gap). Add one line noting these figures reflect only what's visible in the Grasshopper transaction data and should be cross-checked against the processor's own dashboard, since not all fee types may appear identically in bank data.
OUTPUT: Produce exactly these two sections, in this order, with these exact headings:
RULES: If any portion of the prompt is incomplete, please flag it immediately and prompt me back to ensure all portions of the prompt are filled out.
DATA SOURCE: If you're connected to my Grasshopper account(s) via the MCP connector, pull my transaction history directly.
You are acting as my financial analyst.
Pull all transactions from my [ACCOUNT NAME] account for the past [NUMBER, e.g., 12] months.
OUTPUT: Produce exactly these four sections, in this order, with these exact headings:
RULES: Build every figure from the transactions you actually pulled — show monthly totals so I can spot-check them. If fewer than 12 full months are available, tell me and note that the seasonal read is weaker. This is a planning framework based on my past cash flows, not a guarantee of future performance; sweep amounts are starting points for me to adjust. If any portion of the prompt is incomplete, please flag it immediately and prompt me back to ensure all portions of the prompt are filled out
DATA SOURCE: If you're connected to my Grasshopper account(s) via the MCP connector, pull my transaction history directly.
You are acting as my fractional CFO.
My current balance on that account is [CURRENT AMT ON ACCOUNT].
Then:
OUTPUT: First present the analysis in chat as one summary table (Metric | Value | How it was calculated), followed by at most three sentences of commentary. Then generate the PDF described in step 5.
RULES: Use only transactions actually in the account — never estimate or invent line items. Use the most recent transaction date as "today." If any month's data is incomplete, tell me instead of extrapolating. Every figure in the PDF must trace back to a transaction in the account. If any portion of the prompt is incomplete, please flag it immediately and prompt me back to ensure all portions of the prompt are filled out.
DATA SOURCE: If you're connected to my Grasshopper account(s) via the MCP connector, look at my [ACCOUNT NAME/TYPE] account.
You are acting as my bookkeeper preparing a draft expense categorization for my tax preparer to review.
Pull all transactions from my [ACCOUNT NAME] account for [START DATE] to [END DATE].
OUTPUT: Generate a PDF summarizing the content. Produce exactly these three sections, in this order, with these exact headings:
RULES: This is a draft categorization for my tax professional to review — do not present any assignment as a final tax determination, and do not advise me on deductibility. If a transaction could plausibly fit two categories, pick one, mark it with an asterisk, and list the alternative. If any portion of the prompt is incomplete, please flag it immediately and prompt me back to ensure all portions of the prompt are filled out.
DATA SOURCE: If you're connected to my Grasshopper account(s) via the MCP connector, pull my transaction history directly.
Act as an accounting clerk helping me reconcile my Grasshopper statement against my internal ledger.
Present the duplicate flags and round-transfer flags as two separate tables, followed by the reformatted import-ready data. Add one line noting that flagged items are patterns worth reviewing, not confirmed errors, and each should be verified against actual records before any correction is made.
OUTPUT: Produce exactly these three sections, in this order, with these exact headings:
RULES: If any portion of the prompt is incomplete, please flag it immediately and prompt me back to ensure all portions of the prompt are filled out
DATA SOURCE: If you're connected to my Grasshopper account(s) via the MCP connector, pull my transaction history directly.
These prompts are ready, but they access to your live financial data to work. Use our AI connector to securely link your accounts in seconds.