Tools reference · Agents

Create a paper agent

Action create_agent 50 credits a call

Creates an agent that trades your rules on its own paper account (simulated money). It backtests the rules first and returns that result with the new agent.

What it does

Give it the rules from build_rules, or a description and a symbol. The paper account starts with $10,000 and risks 1% of it per trade unless you choose otherwise. The agent starts trading on paper right away unless you ask to save it paused.

The backtest shows the trade count with its results; past results don't predict future ones. Up to 5 agents run at once, and AI apps can create up to 10 agents a day.

A new agent stays on paper: going live needs request_live and your approval in Senntric.

This tool makes a change

It changes something in your Senntric account. It needs the “Make changes” permission when you connect, and your AI app asks you to confirm before it runs. It can’t place an order at your broker.

Inputs

rules Optionalobject. The rules exactly as build_rules returned them. Give rules or a description, not both.
description Optionaltext. Instead of rules, the strategy in your own words. Needs a symbol.
symbol Optionaltext. The ticker or chart symbol. Needed with a description.
name Optionaltext. A short name. The default is the symbol and timeframe, such as "QQQ 5m agent".
risk_usd Optionalpositive number, up to 100,000. Dollars risked per trade on the paper account, from 0.1% to 5% of its size. The default is 1%.
account_size Optionalnumber, 1,000 to 10,000,000. The paper account's starting cash, from $1,000 to $10,000,000. The default is $10,000.
max_trades_per_day Optionalwhole number, 1 to 5. The most trades it takes in a day, from 1 to 5. Leave it out to keep the rules' own limit.
months Optionalone of 3, 6, 12, 24. How much history its first backtest covers: 3 (the default), 6, 12 or 24 months.
start Optionaltrue or false. True (the default) to start trading on paper now, or false to save it paused.

Example

You ask

“Turn those QQQ rules into a paper agent that risks $100 a trade.”

Result (trimmed, example values)
{
  "created": true,
  "agent": {
    "id": "f3a9c2d1-6b7e-4a58-9c0d-2e4f6a8b1c3d",
    "name": "QQQ 5m agent",
    "kind": "rules",
    "symbol": "QQQ",
    "timeframe": "5m",
    "status": "running",
    "version": 1,
    "trades": "paper",
    "live": null,
    "made_by": "Claude",
    "paper_account_usd": 10000,
    "risk_per_trade_usd": 100,
    "max_trades_per_day": 1,
    "open_in_senntric": "https://app.senntric.com/?open=agent&id=f3a9c2d1-6b7e-4a58-9c0d-2e4f6a8b1c3d"
  },
  "backtest": {
    "months": 3,
    "sessions_tested": 62,
    "trades": 41,
    "win_rate": 0.488,
    "avg_r": 0.31,
    "total_r": 12.7,
    "max_drawdown_r": 4.2
  },
  "note": "It is running on paper now. Paper trading: Senntric agents trade simulated money. Going live through TradersPost needs request_live and the user's approval in Senntric."
}

Your AI reads this result and answers in plain words. Most results include an open_in_senntric link to see the same thing in Senntric.

See all tools, example prompts and limits and credits.

Updated November 17, 2026. Education, not investment advice.