Example Conversations

The examples below show the kind of prompts that reliably trigger the FinanceDataAPI MCP surface. They are intentionally explicit so the client has enough context to pick the right tools.

1. Trend check for one ticker

User prompt

Look up AAPL, then compute RSI, MACD, and 20-day plus 50-day SMA from the last 90 days. Summarize momentum, trend direction, and any crossover risk in 2 to 4 short paragraphs.

Likely tool sequence

1. get_symbol
2. get_rsi
3. get_macd
4. get_sma with period=20
5. get_sma with period=50

2. Compare a basket of names

User prompt

Compare AAPL, MSFT, and GOOGL using RSI, MACD, SMA, Bollinger Bands, and ATR from 2025-01-01 to 2025-03-31. Give me a table of the latest values and call out the most overbought and the most volatile name.

Likely tool sequence

1. run_signals for ["RSI", "MACD", "SMA"]
2. run_signals for ["EMA", "BollingerBands", "ATR"]

3. Submit and wait for a backtest

User prompt

Run an RSI mean-reversion backtest for SPY over the last two years with $10,000 starting capital, then wait for completion and summarize total return, Sharpe ratio, max drawdown, win rate, and total trades.

Likely tool sequence

1. submit_backtest
2. wait_for_backtest

4. Use a built-in prompt directly

Some clients let you invoke prompt templates instead of free-form chat. FinanceDataAPI includes three prompt names:

• analyze_ticker
• compare_strategies
• backtest_strategy

For example:

{
  "name": "analyze_ticker",
  "arguments": {
    "symbol": "AAPL",
    "lookback_days": 90
  }
}

Prompt-writing tips

• Mention the ticker symbols, date range, and desired indicators explicitly.
• Ask for a summary, table, or comparison so the client knows to synthesize results after the tool calls finish.
• For backtests, tell the agent to wait for completion instead of polling manually.